Frequently asked questions

WordPress and Next.js are fundamentally different technologies for building websites. WordPress is a CMS (content management system) launched in 2003 that works with templates and plugins, used by about 40 percent of websites worldwide, according to W3Techs. Next.js is a JavaScript framework launched in 2016 by Vercel that allows building custom web applications from scratch.

The practical differences are three. First: speed. A WordPress site with many plugins tends to load more slowly, because every plugin adds code and queries, while a well-built Next.js site serves pages already generated and is usually considerably faster. Second: security. WordPress is a primary target for attacks due to its dominant market share, and outdated plugins are the main entry point. Next.js has a smaller attack surface due to its architecture. Third: maintenance. WordPress requires continuous plugin updates with annual licenses. Next.js has less recurring dependency.

WordPress is suitable for simple content sites with a limited budget. Next.js is suitable for projects that require performance, unique visual identity, or custom functionality that scales.

TypeScript is a programming language created by Microsoft in 2012 that adds static types to JavaScript. All TypeScript code is automatically compiled to standard JavaScript before running in the browser, so end users notice no difference. TypeScript has become the professional standard for modern web development.

The main advantage of TypeScript is detecting errors during development instead of in production. In traditional JavaScript, type errors (using text where a number should go, accessing nonexistent properties of an object) appear when real users run the code. In TypeScript, the editor flags these errors before the file is saved.

The concrete advantages for web projects are three. First: fewer bugs in production, since type errors are the most frequent in JavaScript. Second: more maintainable code in the long term, because TypeScript documents what kind of data each function expects. This makes it easier for a new developer to understand the project. Third: safe refactoring, since TypeScript automatically detects all affected places when a data structure changes.

The cost of TypeScript is somewhat more initial development time, spent writing the types. That time is recovered later, in debugging and maintenance.

Supabase is an open-source Backend-as-a-Service platform launched in 2020 that offers a PostgreSQL database, user authentication, file storage, automatic API, and real-time functions. Firebase is a similar platform from Google launched in 2011.

The fundamental differences are four. First: Supabase uses PostgreSQL (the industry-standard relational database), while Firebase's main database is Firestore, a proprietary NoSQL database (since 2025 Firebase also offers PostgreSQL as an option). Second: Supabase is open source, which allows self-hosting and zero lock-in; Firebase belongs entirely to Google. Third: Supabase costs are predictable (plans starting at USD 25 monthly), while Firebase charges per individual operation and can become unpredictable at scale. Fourth: Supabase supports standard SQL, which makes it easier to hire developers; Firestore requires knowing its own syntax.

Supabase is preferable for applications that may grow in complexity, require advanced queries, or where future data portability matters. Firebase may be suitable for native iOS or Android mobile apps with massive real-time needs where it has a mature track record.

Vercel is a deployment and hosting platform for modern web applications, founded in 2015 by the team that created Next.js. Unlike traditional cPanel hosting, Vercel automatically compiles code uploaded to Git and distributes it globally on a CDN (Content Delivery Network) with servers in multiple regions.

Vercel's main technical features are: automatic deploy from Git, instant rollback to previous versions, preview deployments per change, automatic scaling without configuration, free HTTPS certificates that renew automatically, and globally optimized load times thanks to its CDN.

Vercel's Hobby plan is free but restricted to non-commercial personal use. Commercial projects require the Pro plan starting at USD 20 monthly per developer. Because it serves pages from servers close to the visitor, geographic distance weighs much less on load time than on traditional hosting with a single server.

Vercel is the preferred platform for Next.js sites, React applications, Vue, Astro, and any modern JavaScript framework. It is not suitable for WordPress, which requires traditional cPanel-style hosting.

Core Web Vitals are three technical metrics that Google has used since 2021 to evaluate the user experience of a website. They are a ranking factor in search results and directly affect SEO positioning.

The three metrics are: LCP (Largest Contentful Paint), which measures how long the main visible content of the page takes to appear, with an ideal target of less than 2.5 seconds. INP (Interaction to Next Paint), which measures how long the site takes to respond to the user's interactions throughout the visit (clicks, taps, key presses), with an ideal target of less than 200 milliseconds. CLS (Cumulative Layout Shift), which measures how much content moves while loading, with an ideal target of less than 0.1.

Sites built with Next.js on Vercel start with an advantage on all three metrics, although a heavy site can still fail them. WordPress sites with chained plugins typically fail at LCP due to server response times and at INP due to blocking JavaScript.

Google's official tools to measure them are PageSpeed Insights, which shows data from real users, and the Core Web Vitals report in Search Console, both free. Lighthouse is useful for lab tests, but it does not measure INP: it uses an approximate metric. Optimizing Core Web Vitals is part of modern technical SEO. A slow site loses ranking to faster competitors even if it has better content.

SSR, SSG, and CSR are three different page rendering strategies, each with important implications for performance, SEO, and GEO.

SSR (Server-Side Rendering) generates the HTML on the server every time a user requests the page. It is ideal for dynamic content that changes frequently or depends on the user (dashboards, applications with authentication). The advantage is that the content is complete when it reaches the browser, which benefits SEO and GEO. The disadvantage is that each request consumes server resources.

SSG (Static Site Generation) generates all HTML pages at build time, before users arrive at the site. It is ideal for content that does not change often (blogs, landing pages, corporate sites). The advantage is extreme speed (pages are served from CDN) and low cost. The disadvantage is that updating content requires rebuilding the site.

CSR (Client-Side Rendering) sends JavaScript to the browser that builds the page on the user's device. It was popular between 2015 and 2018 with traditional React. Today it is considered an anti-pattern for public sites because it harms SEO and GEO: crawlers that do not fully execute JavaScript do not see the actual content.

Next.js allows combining all three strategies in the same project as each page requires. This flexibility is one of the reasons for its massive adoption.

The security of a modern website is managed in four layers: infrastructure, code, data, and authentication. Each layer has standard practices that significantly reduce the risk of vulnerabilities.

Infrastructure: modern platforms like Vercel include automatic HTTPS with renewable SSL certificates, basic DDoS protection, and isolation between projects. Unlike traditional hosting, there is no SSH access to the server that could be compromised.

Code: React escapes the content it renders by default, which stops most XSS attacks, and Next.js Server Actions check the origin of every request to block CSRF attacks. SQL injection is prevented in the data layer, with parameterized queries like the ones the Supabase client uses. TypeScript adds a verification layer that reduces errors that could lead to vulnerabilities.

Data: Supabase implements Row Level Security (RLS) at the database level, where access rules are defined directly in PostgreSQL. This means that even if the code has a bug, users cannot access other users' data. Credentials are managed with encrypted environment variables, never written directly into the code.

Authentication: dedicated providers like Supabase Auth, Auth.js (formerly NextAuth), or Auth0 handle login, tokens, refresh, and password recovery with industry-standard practices, preventing the developer from implementing authentication from scratch (a common source of vulnerabilities).

WordPress has additional structural challenges: it is the platform that appears most in reports on infected sites, primarily due to outdated plugins that open attack vectors.

Tailwind CSS is a CSS framework based on utility classes, created by Adam Wathan in 2017 and widely adopted since 2020. It changed the paradigm of writing styles by inverting the traditional logic.

Traditional CSS separates structure (HTML) and presentation (CSS) into different files. The developer invents class names like boton-primario or card-destacada and defines their styles in a separate file. Tailwind inverts this: it offers thousands of predefined utility classes (such as bg-blue-500, px-4, rounded-lg) that combine directly in the HTML.

The advantages of Tailwind are five main ones. First: development speed, since no time is lost thinking up class names or jumping between files. Second: consistency by construction, because the limited values system enforces visual coherence. Third: reduced final file size, since only the classes actually used are included (according to Tailwind, most projects ship less than 10 KB of CSS). Fourth: ease of maintenance, because the style is in the HTML without file jumps. Fifth: full compatibility with traditional CSS when something specific is needed.

The common criticism that Tailwind makes all sites look alike is incorrect. Sites like OpenAI's and Shopify's use Tailwind with completely different visual identities.

GEO (Generative Engine Optimization) is the practice of optimizing web content so it is cited by artificial intelligence engines like ChatGPT, Perplexity, Claude, Gemini, and Google AI Overviews. It was formally defined in 2023 by researchers from Princeton and IIT Delhi.

The difference from traditional SEO is fundamental. SEO optimizes to appear in the top Google results as blue links. GEO optimizes to be included as a source in AI-generated responses. Generative engines synthesize information from multiple sources into a single response, instead of showing a list of links.

The techniques that worked best in that research were citing sources, adding direct quotes from experts and including statistics, along with using precise technical terminology. In its tests, the best ones increased visibility in AI answers by up to 40 percent compared to non-optimized content.

SEO and GEO are complementary, not exclusive. AI engines that search the web need to be able to crawl and index a site, just like a search engine, so good technical SEO is the foundation of GEO.

Yes. Traditional SEO remains critical in 2026 despite the growth of AI search engines. Google still accounts for about 90 percent of search engine use worldwide, according to StatCounter. In 2024, Gartner projected that by 2026 search volume on traditional search engines would drop 25 percent due to the shift toward AI chatbots and assistants.

However, traditional SEO is the foundation that makes AI ranking possible. Generative engines that search the web, like ChatGPT and Perplexity, need to be able to crawl and index a site to use it as a source. A site with deficient technical SEO is at a disadvantage on both sides.

The technical SEO elements that remain fundamental are: load speed (Core Web Vitals), clean URL structure, correct metadata, schema markup, updated sitemap, and indexable relevant content.

The correct strategy in 2026 combines both layers: solid technical SEO as a foundation, complemented by specific optimization for AI engines (GEO) on top of that base.

Schema markup is structured code added to a website's HTML to describe its content in a language that search engines and AI engines understand precisely. It uses a standardized vocabulary developed by Schema.org, a joint initiative of Google, Bing, Yahoo, and Yandex since 2011.

Schema markup explicitly indicates what type of content each page is: a blog article, a product for sale, a local business with address, a recipe, an event, a person, a review with a score. Engines read this code without ambiguity, unlike normal HTML where they must infer meaning from context.

In traditional SEO, schema markup allows Google to display rich snippets in results: review stars, product prices, event times. Google decides which types it shows and changes them over time: expandable frequently asked questions, for example, stopped appearing in May 2026. A richer result can attract more clicks.

In GEO (optimization for AI engines), schema markup helps any system reading the site understand without ambiguity what each page is. It does not guarantee a citation: it removes interpretation work for whoever reads it. The most useful types for a business site are Organization, Article, Product, and FAQPage, which remains valid even though Google no longer shows it as a rich result.

Schema markup is implemented in JSON-LD (the format recommended by Google), embedded in the HTML '<head>'. A professional site can implement multiple types of schema based on the content of each page.

Local SEO is the specific optimization of a business to appear in searches with geographic intent, such as "hairdresser near me", "lawyer in Palermo", or "dentist in Buenos Aires". It differs from traditional SEO by focusing on Google Business Profile (formerly Google My Business) as a central tool, in addition to the website.

A significant share of searches has local intent, and those are the ones closest to a visit or a call. That is why Local SEO is critical for businesses with a physical presence: restaurants, medical offices, professional studios, retail stores, home services.

Local SEO work has five main components. First: create or claim the Google Business Profile listing. Second: complete all the information (hours, phone, address, categories, description, photos). Third: obtain and manage customer reviews, which are a critical local ranking factor. Fourth: verify NAP (Name, Address, Phone) consistency across all platforms where the business appears. Fifth: create content on the website optimized for local searches (location pages, mentions of neighborhood or city).

Local SEO usually delivers results sooner than traditional organic SEO, especially when the Google Business Profile listing was incomplete: improvements in Google Maps and local searches show up within weeks or a few months, while organic ranking usually takes considerably longer.

Google Analytics 4 (GA4) is the web traffic measurement platform launched by Google in October 2020, which replaced Universal Analytics: the standard version stopped processing data on July 1, 2023, and the paid version, Universal Analytics 360, on July 1, 2024. It is the current version and the only one officially supported by Google.

The main differences from the previous Google Analytics are four. First: event-based data model. Universal Analytics measured sessions and page views as central metrics. GA4 treats every interaction (click, scroll, video viewed, form submitted) as an event, allowing more granular analysis of behavior. Second: unified measurement across web and apps. GA4 allows tracking a website and a mobile application in the same property, which Universal did not support. Third: integrated privacy. GA4 is designed to operate without third-party cookies and comply with GDPR. Fourth: native integration with Google Ads and BigQuery.

A correct GA4 setup for a professional site requires approximately 3 to 5 hours. The work includes: installing the tracking code, configuring events specific to the business (WhatsApp clicks, form submissions, purchases), marking the most important ones as key events (what used to be called conversions), filtering internal team traffic, and verifying with real tests.

An alternative to GA4 is Plausible Analytics, a paid option (USD 9 to 19 monthly) with a simpler interface, no cookies, and no data sent to Google. Plausible is suitable for sites that prioritize user privacy.

Preparing a site to be cited by AI engines requires six technical and content practices, based on the Princeton and IIT Delhi research on GEO and on the engines' own documentation.

First: allow AI crawlers access in the robots.txt file. Each company splits its bots by purpose. To appear in answers that use search, the search bots are what matter: OAI-SearchBot (OpenAI), Claude-SearchBot (Anthropic), and PerplexityBot. Training bots, such as GPTBot and ClaudeBot, and the Google-Extended token are decided separately, and blocking them does not remove the site from those searches. If the site uses Cloudflare, check its settings: since 2025 it blocks some of these bots by default on new domains.

Second: implement JSON-LD schema markup on important pages, especially Article, FAQPage, and Organization. Structured data tells any system without ambiguity what type of content each page is.

Third: answer-first structure on each page. The start of each page must directly answer the main question, not build toward the answer. It is the first thing a system summarizing the page finds.

Fourth: data density with named sources. Specific statistics, concrete dates, and references to authoritative sources (Gartner, HubSpot, academic research) were, in the Princeton research, among the techniques that increased visibility the most, by up to 40 percent.

Fifth: up-to-date content. An outdated fact is a reason not to cite a page, and a visible update date, along with the dateModified field in the schema, says when it was last reviewed. That date has to be real: changing it without reviewing the content does not make it current.

Sixth: presence in the communities where the industry is discussed, such as Reddit. Genuinely answering questions, without promotion, generates indirect authority.

Tools to measure citations in AI engines are a new segment that emerged during 2024 and 2025. There are currently four main options with differences in price and depth.

Profound (tryprofound.com) is among the most complete. It tracks brand mentions in ChatGPT, Claude, Perplexity, Gemini, and Google AI Overviews, and allows comparing with competitors and measuring share of model (a brand's share versus the competition in AI answers for the industry). It now targets companies and quotes custom pricing.

OtterlyAI (otterly.ai) specializes in monitoring Google AI Overviews and ChatGPT. It is simpler than Profound but sufficient for most businesses just starting to measure. Price starting at USD 29 monthly.

Peec (peec.ai) is another specialized option, with plans from 89 euros monthly and a trial period.

Manual tracking is the zero-cost option to start. It consists of asking ChatGPT, Perplexity, and Google AI Overviews 10 to 20 queries related to the business once a month, and recording in which queries the site appears cited. It works as a baseline but scales poorly.

In June 2026 Google Search Console began showing, for some sites, performance reports for AI Overviews, AI Mode, and Discover. For now they report impressions, not clicks. They are free and worth reviewing regularly.

For a business just starting with GEO, the recommendation is to combine monthly manual tracking with a review of Google Search Console. As content volume grows, it is worth evaluating a paid tool like OtterlyAI or Peec.

There is no single price: a professional website in 2026 can cost anywhere from a few hundred to tens of thousands of dollars, depending on the type of site, the required functionality, and the level of customization. There is no reliable public survey of website prices, so the figures in this answer are the reference Senda Lógica uses when comparing proposals, not a market study.

Simple landing pages (a single page with contact form and custom design) range from USD 400 to USD 1,200. Corporate or service sites with 5 to 8 pages, blog, and basic SEO optimization range from USD 1,200 to USD 3,000. Ecommerce sites with complete online store, payment integration, and inventory management range from USD 3,000 to USD 8,000. Web applications with authenticated users, dashboards, and custom business logic start at USD 6,000 and can exceed USD 20,000 in complex projects.

The factors that most affect the final price are: number of unique pages or sections, functionality complexity (ecommerce vs informational site), need for custom design vs use of templates, integrations with external systems, and level of visual customization.

By who does the work, the same reference places a professional freelancer between USD 500 and USD 4,000, a boutique studio between USD 1,500 and USD 8,000, and a traditional agency from about USD 5,000, with large projects going well beyond that.

Price should be evaluated alongside scope, not in isolation. A quote without clearly specifying what is included is a risk signal.

Senda Lógica publishes its prices and deliberately sits at the low end of those references, and below them for the store and custom work: Essential Page USD 600, Landing Page USD 1,200, Online Store USD 2,500, and Custom Project from USD 4,000. The difference is not in cutting scope but in how the work is built: AI-assisted development on a modern stack, with no server configuration hours and no plugin licences to pass on to the budget.

A professional website has recurring costs that vary significantly between a personal or experimental project versus an active commercial project. The distinction is important for proper planning.

For personal and experimental projects, with no commercial operation at all, costs can be close to zero. Vercel offers a free Hobby plan with 100 GB of monthly bandwidth and up to 100 deployments per day, sufficient for most personal sites. Supabase offers a Free plan with 2 projects, 500 MB of database, and up to 50,000 monthly active users. The only relevant limitation of Supabase's Free plan is that projects inactive for more than 7 days are automatically paused and require manual reactivation. For a personal site with sporadic traffic this is not a problem.

For professional commercial projects the recommendations are different. Vercel Pro costs USD 20 monthly per developer and explicitly enables commercial use, which is technically not permitted on the Hobby plan according to the terms of service. Supabase Pro costs USD 25 monthly and eliminates the risk of automatic pauses, guaranteeing continuous availability for sites serving real clients.

Other common recurring costs include the domain (between USD 10 and 20 annually), corporate email with Google Workspace (from about USD 7 monthly per user), and optional services such as email marketing or headless CMS.

On the sites Senda Lógica delivers these costs are published per package: between USD 20 and 40 monthly for Essential Page and Landing Page, which run on a Pro plan because the free one does not allow commercial use; between USD 45 and 95 monthly for Online Store, which adds a database and depends on traffic; and between USD 10 and 20 annually for the domain. All of it stays in the client's name and is paid directly to the provider, with no intermediaries and no studio markup.

The honest recommendation depends on the state of the project. Newly launched personal sites with little traffic can start on free plans and migrate to paid plans when the project justifies it. Established commercial projects with real traffic should operate on paid plans from the start to avoid availability risks and comply with the platforms' terms of service.

The price of a professional web project is determined by five structural factors that act in combination. Understanding these factors allows fairly evaluating comparable quotes and detecting proposals that underestimate the actual scope of work.

Factor one: project scope. The number of unique pages, the complexity of required functionalities (forms, integrations, private areas, ecommerce), and the level of design customization determine the total work hours. A project with custom design requires considerably more hours than one built on templates.

Factor two: technology used. Development with modern frameworks like Next.js requires more initial investment than configuring WordPress templates, but produces faster, more secure sites that are easier to maintain in the long term. The technology choice affects both the initial price and future operating costs.

Factor three: developer's professional experience. A professional with established experience charges more per hour than someone in training, but makes better architectural decisions that impact the site for years. As the studio's reference, the hourly rate for professional web development in Latin America usually runs from USD 20 to USD 80 depending on experience and specialization.

Factor four: services included in the proposal. Complete professional projects include initial diagnostic, visual design, formal review rounds, testing on multiple devices, technical SEO, client training, documentation, and post-delivery warranty. Cheaper proposals usually limit these services or do not include them.

Factor five: infrastructure ownership. Projects where the client owns their hosting, domain, and service accounts from the start require more configuration work than projects where everything stays in the developer's account. The first option protects the client in the long term; the second creates dependency.

The practical recommendation when evaluating quotes is to request detailed scope in writing before comparing prices.

Freelancer, boutique studio, and agency are three different work models with important differences in price, process, and client experience. The right choice depends on the type of project and the client's needs.

A freelancer is an independent professional working alone, usually taking projects one at a time. Advantages: direct dealing, more accessible prices (as the studio's reference, between USD 500 and 4,000 for a professional web project), process flexibility, zero bureaucracy. Disadvantages: limited capacity, total dependency on one person, limited availability outside business hours.

A boutique studio is a small structure — from a single person up to eight — with clear specialization and a formal process. Advantages: scope and price defined in writing before starting, deep specialization in a given stack, and direct dealing with whoever builds. Disadvantages: limited capacity compared to an agency. Price varies a lot by model: a studio working with a modern stack and AI-assisted development sits well below one quoting by traditional hourly rates, even when the deliverable is equivalent.

A traditional agency is a structure of 15 or more people with multiple departments. Advantages: capacity for large and complex projects, broad multidisciplinarity, corporate processes. Disadvantages: high prices (from about USD 5,000, and considerably more on large projects), indirect communication through project managers, more bureaucracy, high staff turnover that affects continuity.

For SMBs, entrepreneurs, and independent professionals, freelancers or boutique studios tend to be the best option for price-quality ratio and direct dealing. For large corporate projects with multiple stakeholders, traditional agencies justify their cost through management capacity.

Senda Lógica is a one-person studio that takes projects end to end without outsourcing, with its prices published on the site. Being able to see the price before the first conversation is itself information about the model: the scope of each package is written down, and so is what falls outside it.

Correctly quoting a web development project follows a 5-step process that protects both the client and the developer. Skipping this process usually results in cost overruns, disputes, or deliverables that do not meet expectations.

Step one: initial diagnostic conversation at no cost. The professional developer asks questions about the business, objectives, audience, references of sites the client likes, approximate budget, and expected timeline. This conversation takes 30 to 60 minutes.

Step two: written scope definition. With the diagnostic information, what the project specifically includes is documented: number of sections or pages, required functionality, integrations, approval points, timeline, and what is NOT included (important). This document is shared with the client for approval before moving forward.

Step three: numerical quote with breakdown. Total price, suggested payment schedule (typically 50 percent upfront, 50 percent on delivery; or 40-30-30 by milestones), estimated delivery time, and special conditions.

Step four: negotiation of adjustments if applicable. The client may want to remove functionality to reduce price, or add something not anticipated. These conversations before the start are healthy and normal.

Step five: formal agreement before starting. It can be a formal contract or confirmation email depending on the type of relationship, but a written record is kept of the final agreed scope and conditions.

A valid shortcut for the client is to start from published prices. When a studio publishes the price and scope of each package, steps one through three get shorter: the conversation stops being about how much it costs and becomes about which one fits.

Risk signals when quoting: prices given without knowing the project, promises without detailed scope, absence of clear timeline, or pressure to decide quickly without time to evaluate.

Generative AI tools for web development like v0 (Vercel), Bolt (StackBlitz), Lovable (formerly GPT Engineer), and Replit Agent allow generating complete websites with natural language prompts. They are useful for specific cases but have important limitations in 2026.

AI-first tools work well for: rapid prototypes to validate ideas, simple landing pages for personal use, temporary sites (events, promotions), and design experimentation without commitment. In these cases they save most of the time compared to traditional development.

Real limitations appear when the project moves from experimental to productive. First: quality of the generated code. AI produces code that works but usually has performance, accessibility, and scalability issues that require deep refactoring before production. Second: maintainability. A developer who did not write the code has difficulty modifying it later. AI-generated code lacks the architectural coherence that an experienced human provides. Third: debugging. When something fails in production, someone has to understand why. If nobody reviewed the code when it was generated, that diagnosis starts from zero. Fourth: technical limitations. Complex integrations, specific business logic, and advanced optimizations are usually beyond the reach of these tools.

The correct professional use combines both worlds: a human developer makes architectural decisions, uses AI as an assistant to speed up mechanical parts of the code, and reviews or adjusts what is generated. This approach shortens development without giving up professional quality.

For a business that needs its site as a serious asset (real traffic, online sales, customer base), professional development assisted by AI remains the most solid option. For experimental or temporary projects, AI-first tools are sufficient.

Mercado Pago is the payment processing platform of Mercado Libre, launched in 2003 and available in Argentina, Brazil, Mexico, Chile, Colombia, Peru, Uruguay, and Ecuador. It processes payments with credit and debit cards, bank transfers, cash at payment points, and the balance of the Mercado Pago virtual wallet.

The technical integration of Mercado Pago into a custom website requires approximately 4 to 6 hours of professional work. The process includes: configuring the seller account to obtain credentials (Access Token and Public Key), writing server code that creates payment preferences via API, configuring return URLs for approved, rejected, and pending cases, implementing webhook handling that receives automatic notifications when a payment is completed, and testing all scenarios with test cards before activating production.

Mercado Pago offers basic integration at no cost on its end. The client only pays commissions on completed transactions (variable based on payment method and country). The developer's cost corresponds to the technical work of connecting the site with the Mercado Pago API, not to the platform itself.

WhatsApp Business and WhatsApp Business API are two distinct Meta products with similar names. WhatsApp Business is a free mobile application for cell phones, designed for small businesses serving from a single phone. It includes features such as product catalog, automatic replies, business hours, and contact labels. It is downloaded from Google Play or App Store.

WhatsApp Business API is a completely different enterprise platform, aimed at companies with high message volume or teams that need to manage support across multiple agents. It allows connecting WhatsApp with websites, CRMs, and proprietary systems, implementing automated chatbots, sending massive notifications with pre-approved templates, and having multiple users handling the same business account.

The API, which Meta now calls the WhatsApp Business Platform, is paid and requires a verification process with Meta. There are two costs. Meta charges for each template message delivered (marketing, utility, or authentication), with rates that depend on the category and the country; replies within the 24-hour customer service window have their own conditions, which Meta keeps adjusting. And if an access provider is used, such as Twilio, 360dialog, or Zenvia, its fee is added: some charge per message and others a monthly fee per number. The Cloud API can also be used directly with Meta, with no intermediary.

For most small and medium businesses, the free WhatsApp Business app is sufficient. The API is justified in operations with high consultation volume, multiple support agents, or need for advanced automations.

Integrating an email marketing system like Mailchimp or Brevo into a professional website combines two costs: the initial setup work, quoted case by case, and the monthly cost of the chosen platform, which the client pays directly to the provider.

Mailchimp offers a free plan for up to 250 contacts and 500 sends a month, with paid plans starting at USD 13 monthly. Brevo (formerly Sendinblue) offers a free plan of 300 daily sends and up to 100,000 contacts, with paid plans starting at approximately USD 9 monthly. Mailchimp tends to be more suitable for beginners due to its visual interface, while Brevo is better for large lists or integration with SMS and WhatsApp.

The professional setup work takes around 4 to 5 hours and includes: account creation and verification with custom domain to avoid spam, main list configuration with custom fields, integration of the subscription form to the site via API or embed, design of the automatic welcome email, configuration of the automated email sequence, legal compliance with GDPR and CAN-SPAM (consent checkboxes, unsubscribe links), and client training.

Email marketing remains one of the channels with the best return: in Litmus's survey, the return most companies report is between USD 10 and USD 36 for every dollar invested.

A headless CMS is a content management system that separates the administration panel from the public site. Traditional CMSs like WordPress operate with a database, panel, and public site coupled together. A headless CMS like Sanity Studio, Contentful, Strapi, or Payload keeps only the administration panel and delivers content via API to the site, which is built with another technology.

The advantages of a headless CMS are three main ones. First: technological flexibility, since the public site can be built with Next.js, React, Vue, or another modern technology without depending on the CMS. Second: better performance, since the public site can be static and ultra-fast. Third: better editing experience, with modern Notion-style interfaces instead of old administration panels.

A headless CMS is suitable when the client needs to edit content frequently and with its own structure: a blog with several authors, a catalogue with changing fields, new content types. It is not suitable for static sites that rarely change, where it adds unnecessary complexity.

One point worth clearing up, because it is often confused: editing your own content does not always require a headless CMS. At Senda Lógica, Landing Page and Online Store already include an editing panel in the package price — text and images of the existing sections, and in the store also adding and removing products with no cap. A headless CMS only enters the picture when what is needed goes beyond that.

An API (Application Programming Interface) is an interface that allows two software systems to communicate with each other by exchanging data in a structured format, typically JSON. APIs are the foundation of practically all modern web applications.

The most common APIs are REST (Representational State Transfer), which use URLs and standard HTTP verbs (GET to read, POST to create, PUT to update, DELETE to delete). GraphQL is a more modern alternative that allows the client to request exactly the data needed in a single query.

In the context of a professional website, APIs are used to connect the site with external services. For example: Mercado Pago has an API to process payments, Google Maps has an API to display maps, WhatsApp Business API allows automating messages, and management systems like Contabilium or HubSpot have APIs to synchronize data.

Integrating third-party APIs into a site requires specific development work: authentication with credentials (API keys, OAuth tokens), proper handling of responses and errors, respect for usage limits (rate limiting), and exhaustive testing before activating in production.

The cost of integrating an external API varies considerably. Well-documented and modern APIs (Mercado Pago, HubSpot, Contabilium) take between 4 and 25 hours. Old or poorly documented APIs (proprietary ERP systems) can take 30 to 80 hours or more.

The most common integrations on professional Latin American sites reflect the region's particular ecosystem, different from the United States or Europe due to predominant use of certain local platforms.

Online payments: Mercado Pago is the most widespread method in Argentina and has a strong presence in Brazil, Mexico, Chile, Colombia, Peru, Uruguay, and Ecuador, although in Brazil it competes with Pix and in Chile with Webpay. It integrates payments with cards, transfers, cash at payment points, and virtual wallet. Stripe is an alternative for projects oriented to the international market.

Messaging: WhatsApp is the dominant messaging channel in almost all of Latin America. Basic integration as a contact button is standard on almost any professional site. WhatsApp Business API for complex operations exists but is less common due to costs.

Business management: popular ERPs vary by country. In Argentina: Tango Gestión (medium-sized companies), Contabilium (modern SMBs), Xubio, Colppy, Bejerman. In Brazil: Omie, ContaAzul, Bling. In Mexico: Contpaqi, Aspel. Integrations with these systems are frequent in B2B ecommerce projects.

Email marketing: Mailchimp and Brevo are the most widely used global options.

CRM: HubSpot dominates the SMB segment due to its robust free plan. Zoho is popular in the small segment. Salesforce maintains its position in large enterprises.

Analytics: Google Analytics 4 remains the dominant standard. Plausible grows in privacy-conscious niches.

Logistics (ecommerce only): Andreani and OCA in Argentina, Correos de México, Coordinadora and Servientrega in Colombia, Chilexpress in Chile. Integrations require specific development based on country and company.

The most frequent integrations in Latin American ecommerce projects are: Mercado Pago plus WhatsApp plus Google Analytics 4 plus basic email marketing. That combination covers most SMB cases.

The work process at Senda Lógica has 5 steps, each with a concrete deliverable and a client approval point before moving to the next.

Step 1. We talk. The client explains the business and what they need to achieve. Questions come before proposals: no diagnostic, no proposal. It is a 30 to 60 minute conversation, at no cost and with no commitment.

Step 2. The proposal arrives. Scope, price, and timeline in writing, with what is included and what is not listed out before deciding. When the project fits a published package, this step is short: the price and the scope are already on the site.

Step 3. The content is prepared. The client is asked for the only thing the studio cannot supply: text, images, logo, and which sections they want. A clear list is provided so it does not become an open-ended task. If there is no text or imagery, that gets solved during this step. It is the step that most often delays a project, and the only one that depends on the client.

Step 4. Design and build. The site is built with the client's identity, showing progress along the way and adjusting with their input before launch. Adjustments happen here, on work that has not gone to production yet.

Step 5. Launch and support. The site goes live on its final domain, measurement is configured, and the 3 months of post-delivery support begin. The code and the infrastructure belong to the client from day one, not from delivery.

Total timeline by package, counted from the moment the content has been handed over: Essential Page 2 to 4 business days, Landing Page 3 to 7 business days, Online Store 3 to 5 weeks. Custom Project is defined in writing in the proposal.

These times reflect modern AI-assisted development, which significantly reduces mechanical work compared to traditional development without lowering the quality of the deliverable. What does not speed up is the back and forth of content and approvals, which is where most of the calendar goes.

The developer needs specific information from the client in 4 categories to execute a project efficiently. Preparing these elements speeds up the process and reduces iteration costs.

Category 1. Business information. Clear description of the business, unique value proposition, target audience, direct competitors with URLs, specific objectives for the site (sell, generate enquiries, inform), and how success will be measured (enquiries per month, monthly sales).

Category 2. Site content. Text for each section (even initial drafts), product or service images in good quality and original format, vectorized business logo (SVG or AI preferred), color palette if the business has a defined visual identity, testimonials and cases with permission to use, and relevant legal documents.

Category 3. Technical access if applicable. Domain account where the site will live, accounts for existing services to be integrated (Mercado Pago, email marketing, current CRM), administrative access to the current site if migrating, and social media credentials if they will be integrated.

Category 4. References and preferences. Competitor or industry sites the client likes, and what specifically they like about them; sites they explicitly do NOT like, which is equally useful for avoiding wrong directions; visual style preferences; and concrete constraints such as mandatory corporate colors.

Many clients do not have all of this ready at the start, and that is normal. In Senda Lógica's process this request does not arrive as an open-ended list: it is step 3, with a guide of what is needed and in what format, and if text or images are missing that gets solved right there. Having as much prepared as possible beforehand shortens the project and avoids blockers mid-development.

Changes during a web project are managed with per-stage approval points and clear agreements about what constitutes "change within scope" versus "change out of scope". This system protects both the client and the developer.

How adjustments are ordered: each stage closes with an approval before moving to the next, and changes are requested against the progress shown, not against an already published site. The first adjustments tend to be structural (layout, information hierarchy) and the last ones fine-grained (typography, spacing, color). Asking for them consolidated by stage, rather than as they occur to you, is what avoids redoing work.

Changes within scope include: content adjustments (text, images), minor visual refinements (color shades, sizes), modifications within the approved structure. These generate no additional cost.

Changes out of scope include: adding new unplanned sections, changing the project's base technology, completely redesigning an already approved section, adding complex new functionality. These require additional quoting and may extend the timeline.

When the client requests a change, the developer must evaluate and communicate clearly: whether it is within or outside scope, whether it affects the timeline, and whether additional cost applies. This communication should be in writing to avoid misunderstandings.

The healthy practice is for minor changes within scope to be implemented without friction, and major ones discussed openly. An accumulation of major changes during a project usually signals that the initial planning stage was insufficient, not that the client is difficult.

If the initial design does not meet expectations, there is a structured process to realign before moving forward. The professional process is designed specifically so this situation involves no additional cost or conflict.

Step one: guided design review. Instead of "I like it" or "I do not like it", the professional developer asks specific questions: what works well, what does not, which references inspire the client most, what emotion they want to convey. This turns vague feedback into actionable direction.

Step two: root cause identification. There are three typical causes of design rejection. First: disconnect with the brand, when the design does not reflect how the client sees their business. Second: disconnect with the audience, when the design looks professional but does not resonate with the target. Third: subjective personal preferences of the client. Each cause requires different treatment.

Step three: adjustments within the normal process. Adjustments at this stage are part of the work, not an extra. Changes to color palette, typography, image style, or visual distribution are expected while the design is still under review, and that is exactly why it is shown in the progress previews, before anything is published.

Step four: if fundamental disagreement persists, pause and rethink. On rare occasions the client's vision and the developer's proposal do not converge. In that case the honest conversation is: return to the scope document and clarify what is being sought, or mutually acknowledge that the match was not right and part ways with what has been delivered.

Step five: prevent this situation in future projects. A well-run diagnostic stage — references, sites that are disliked, style definition — greatly reduces the chance that the first design misses expectations. A well-managed project reaches the final design with clear client approval, without needing to go back.

After formal delivery of the site, the relationship with the developer continues in three distinct ways depending on the agreed model. Clarity about which model applies is part of the initial quote.

Model 1. Included post-delivery support. At Senda Lógica every package includes 3 months at no additional cost, and the scope is published: unlimited technical service for bugs that appear in production, up to 3 minor content adjustments (text, images, contact details), and verification that everything works correctly in production. It does not cover structural changes to the delivered site. This period is not long-term recurring maintenance, but the professional warranty on the work delivered.

Model 2. Continuous monthly maintenance. Once the support period ends, the client can take the Care Plan: security updates and patches, backup checks, uptime monitoring and one hour of minor changes per month, with a short report. Updates don't count against that hour. It is optional, with a minimum 3-month commitment. If the site has the Consultation Agent, there is also the Agent Plan, which covers its AI usage and the review of its conversations, and the Complete Plan, which combines both.

Model 3. On-demand work. For sporadic needs there are hour banks, usable within a set period, and one-off changes at a fixed price each: small ones are text, contact details, and adjustments that do not alter the structure; medium ones are a form or reorganizing a layout. All of these options are listed on the complementary services page with their price published.

What none of these models includes: major new functionality, site redesign, fundamental structural changes. Those are new projects requiring specific quotes.

What the client should always have: access to their own hosting, domain, database, and analytics accounts. This guarantees that if they eventually change developer, the transition is simple. At Senda Lógica, with the one-off payment, those accounts are created in the client's name from day one, not at delivery; with the base plan's monthly fee, hosting lives on Senda Lógica's account and the domain stays in the client's name.

Continuity with the original developer tends to be beneficial because they know the project in depth. However, a site built with standard technologies and good practices allows any capable developer to take over maintenance if necessary.

A slow website typically has 5 root causes. Identifying the specific cause is crucial before applying solutions, because optimizing the wrong aspect does not improve actual speed.

Cause 1. Heavy unoptimized images. It is the most common cause, especially on photography sites, ecommerce, and blogs. Images in old formats (unoptimized JPG, PNG) and massive sizes can add whole seconds of loading. Solution: conversion to WebP or AVIF, compression with tools like Squoosh, responsive sizes appropriate to each device.

Cause 2. Slow or inadequate hosting. Cheap shared hosting can have saturated servers and high response times. Solution: migrate to modern platforms like Vercel (for Next.js sites) or upgrade traditional hosting if it is WordPress.

Cause 3. Excessive plugins (WordPress sites). Each plugin adds code that runs on every load. With dozens of active plugins, loading can become very slow. Solution: plugin audit, eliminate unused ones, replace multiple small plugins with more comprehensive solutions.

Cause 4. Blocking JavaScript. Scripts that load before visual content delay the Largest Contentful Paint. Solution: lazy loading, async or defer scripts, eliminate unused JavaScript.

Cause 5. Lack of CDN or caching. Serving content from a single geographic server slows down users in other regions. Solution: implement CDN (modern platforms like Vercel include it automatically).

Free diagnostic tools: Google's PageSpeed Insights, Lighthouse (integrated in Chrome DevTools), GTmetrix. PageSpeed Insights shows all three Core Web Vitals (LCP, INP, and CLS) with data from real users, when the site has enough traffic; Lighthouse and GTmetrix are lab tests and replace INP with an approximate metric. All three give specific recommendations.

The reference is Google's: the main content (LCP) should appear in 2.5 seconds or less, the response to each interaction (INP) should take 200 milliseconds or less, and content should move little while loading (CLS of 0.1 or less). Above those values the experience suffers, and Google takes it into account.

An absent developer situation is stressful but has a structured solution in 5 steps. Prevention is better than resolution, but if it has already happened, there are clear paths to regain control.

Step 1. Assess how serious the situation is. If the site is still working and you only need minor changes, urgency is low. If the site is down, the domain expires soon, or there is important information only the developer has, urgency is high. This evaluation defines subsequent actions.

Step 2. Inventory what accesses you have. Check if you have direct access to: domain account, hosting account, corporate email account, social media account, source code files, database. Each access you have significantly reduces risk.

Step 3. Try orderly recovery before escalation. Contact the developer through multiple channels (email, WhatsApp, LinkedIn, mutual acquaintances) with a clear message about specific need and deadline. A formal email leaving written record is useful.

Step 4. Transfer ownership of accounts you do not have. If the domain is in the developer's account, contact the registrar (GoDaddy, Namecheap or, for .ar domains, NIC Argentina) with evidence of business ownership to start the transfer process. Timelines depend on the registrar and the paperwork, so it is best to start as soon as possible. If the hosting is in the developer's account, the entire site has to be migrated to your own account, which requires hiring a new developer.

Step 5. Hire a new developer for audit and recovery. A professional can: access the code if it is on shared GitHub, extract the current site's content if it still works, recreate functionality on new infrastructure controlled by the client. As the studio's reference, a recovery usually costs between USD 500 and 3,000 depending on complexity.

Prevention for the future: always insist that all infrastructure be in client accounts from the start of the project. The developer must be a collaborator, never an owner. This practice avoids the problem completely.

A Google Ads investment that does not convert typically has one of 4 causes, and the root cause is usually on the site, not in the campaign. Before increasing budget or changing agency, it must be diagnosed correctly.

Cause 1. Problems in the Ads campaign. Keywords too generic or competitive, deficient targeting, ads with low relevance, insufficient budget to be competitive. Solution: campaign audit by a performance marketing professional. It is detected through analysis of impressions, CTR, and quality score.

Cause 2. Landing page misaligned with the ad. When the user clicks on an ad for "online pizza order" and arrives at the restaurant's general homepage, friction is high. Solution: campaign-specific landing pages that respect the ad's message exactly. A dedicated landing page, which continues exactly what the ad promised, usually converts better than sending people to the homepage.

Cause 3. Site not optimized for conversion. The user arrives but the site does not guide them toward the expected action. Common problems: CTA not visible above the fold, forms too long, lack of trust elements (testimonials, certifications), slow site that they abandon before converting. This is the most frequent cause and the most ignored.

Cause 4. Weak offer or misaligned prices. Sometimes the problem is not technical but product-market: the price is not competitive, the offer is not attractive, or the target audience is incorrect. Diagnosis: compare with competitors who do convert. Solution: offer adjustment, not site adjustment.

The correct order to optimize is: first the site (ensure it converts), then specific landing pages, then the Ads campaign. Spending more on Ads with a deficient site is amplifying the problem.

There is no reliable regional conversion-rate benchmark. The best reference is your own rate before and after each change, measured on the same kind of traffic: if it drops or does not move, the problem is in one of the 4 causes.

An old site has three possible paths depending on the technical diagnosis and the business objectives. The right decision depends on honestly evaluating the cost-benefit of each option.

Option 1. Update and maintain the existing site. Applies when: the base technology is still reasonable, the design is not that outdated, the content is still relevant, and there are no structural performance or security problems. Typical work: update plugins and themes, optimize images, adjust design in specific sections, improve technical SEO. At Senda Lógica this is handled with one-off services contracted separately, without rebuilding the site.

Option 2. Refactor while keeping content. Applies when: the site structure is still valid but the base technology is obsolete, the design needs major renovation, or performance is unacceptable. Work: rebuild the site on modern technology while keeping content, basic structure, and URLs so rankings are not lost. Migration from WordPress, Wix, TiendaNube, or others is quoted according to the volume of content to move and how structured it is at the source.

Option 3. Rebuild completely from scratch. Applies when: the business changed significantly, the current site is misaligned with the brand's present identity, there are serious architecture problems, or 5 or more years have passed and everything needs renewal. Work: a complete new project, with careful migration of rankings and important data. The cost is that of the corresponding package, which at Senda Lógica starts at USD 600 and is published.

Clear signals that it is time to act: loading in more than 4 seconds, failures on mobile, inability to edit content, visible errors in sections, an appearance similar to sites from 2010, sustained decline in organic traffic.

Before deciding, a technical audit is worthwhile: it diagnoses the current state and returns prioritized recommendations. It is a small investment next to the project, and it avoids spending a lot on the wrong option.

Progressive loss of Google ranking is common but not inevitable. Understanding the causes allows distinguishing between normal losses (recoverable) and symptoms of bigger problems.

Cause 1. Google algorithm updates. Google releases several core updates per year and thousands of smaller changes: more than 4,000 in a year, according to Google itself. Some sites lose positions after these updates for reasons that are not entirely transparent. Solution: wait at least a week after the rollout finishes before drawing conclusions, as Google recommends, and keep in mind that a recovery can take months; review the official guides of the update, adjust content if specific problems are detected.

Cause 2. Outdated content. Google favors relevant and updated content. Articles from 3 or more years ago with obsolete information lose ranking against competitors who publish new content. Solution: quarterly update program for important articles, addition of visible "Last updated" date, refresh of statistics and examples.

Cause 3. Growing competition. New competitors enter the market with better SEO, more content, or better technical optimization. The relative position of the site drops even though the site itself has not worsened. Solution: periodic competitive analysis, identify what successful competitors are doing, invest in areas where you are falling behind.

Cause 4. Accumulated technical problems. Outdated plugins, broken internal links, error pages, degraded speed, expired SSL certificate. Solution: semi-annual professional technical audit, correction of detected errors.

Cause 5. Change in search intent. The same keywords can change intent (users search for something different with the same terms). The site can become misaligned. Solution: Google Search Console monitoring to detect changes, content adjustment to new intent.

Effective prevention: monthly position review in Search Console, regular update of important content, Core Web Vitals monitoring, and proactive adjustments at early signs of decline. SEO is continuous maintenance, not a one-time project.

A website for a healthcare professional (doctor, psychologist, nutritionist, physiotherapist, dentist) has specific requirements combining professional information, patient acquisition, and operational clarity.

Essential content: professional biography with verifiable training and credentials, specialties with a clear description of which problems are treated, practical information (address, hours, accepted insurance, modes of care), an effective contact channel — WhatsApp is usually preferred — and cancellation and first-visit policies.

What is worth solving on the site and what is not. Presenting the practice, organizing the information, and making contact easy is what pays off from day one. Anything that involves receiving clinical data, however, deserves more care than it usually gets: a form collecting health information creates obligations over that data, and a private patient area with medical records is a development with its own legal requirements, not just another section. Senda Lógica does not include them in its packages for that reason: they are discussed as a separate project, with their own scope, or solved with existing scheduling tools such as Doctoralia.

Industry-specific considerations: compliance with local healthcare advertising regulations, which vary by jurisdiction; careful handling of patient testimonials, with written consent and anonymization where appropriate; avoiding unsupported medical claims; and clear information about privacy and data protection.

Critical SEO for healthcare professionals: Local SEO is the priority because most patients search for "[specialty] in [area]". A well-configured Google Business profile produces results quickly. Reviews from real patients weigh heavily on the decision. Educational content about the conditions treated attracts patients in the research stage.

Recommended design: restrained and professional, conveying trust. Avoid aggressive aesthetics or pushy marketing looks. Professional photos of the practice and the professional build confidence.

What to invest: at Senda Lógica a healthcare professional enters through Essential Page (USD 600) when what they need is to exist properly and be reachable, or through Landing Page (USD 1,200) when they have several specialties to explain, recurring questions, and want the site to bring in enquiries. Charging for registrations, integrating scheduling, or adding a private area are add-ons quoted separately case by case.

A website for a law firm combines professional positioning, client acquisition, and building legal authority. Its requirements differ from other professional sectors.

Essential content: presentation of the firm with its history and values, biography of each lawyer with training, experience, and specialties, practice areas describing which cases each one handles, location, and contact methods with office hours.

Differentiating content: legal analysis of current topics, highly valued by both search engines and AI engines; anonymized cases with measurable results, respecting professional privilege; academic publications; media appearances; and frequently asked questions specific to each practice area.

Useful functionality: a first-consultation form with questions specific to each area, and scheduling if the firm already works that way. A private client area with case documents is a separate development with its own confidentiality requirements: it does not fit in a standard package and is best treated as an independent project.

Industry considerations: respect for the professional advertising rules of each bar association, which vary by jurisdiction; caution with claims about success rates, which may be considered misleading advertising; appropriate handling of confidential information in testimonials; and avoiding specific legal advice in public content.

Critical SEO for law firms: authority content is fundamental. Lawyers who publish serious analysis in their specialty rank well and get cited by AI engines. Local SEO matters if the firm serves a defined area. Competition on generic keywords like "divorce lawyer" is high; focusing on specific niches works better.

Recommended design: serious and restrained, conveying authority. Classic typography, a contained palette, professional photography of the lawyers and the office.

What to invest: at Senda Lógica a law firm enters through Landing Page (USD 1,200), the package built for services that live on enquiries: up to 8 sections to present practice areas, partners, and cases, plus frequently asked questions. If a multi-author blog or a client area is needed later, that is a Custom Project and is quoted by scope.

A small ecommerce in Latin America has specific regional market considerations affecting technology, logistics, and monetization decisions. The right strategy balances initial investment with capacity to grow.

Fundamental decision 1. Platform versus your own store. Platforms like TiendaNube or Shopify let you start quickly at USD 20 to 80 monthly in subscription plus commissions, in exchange for a store that looks like the others on the same platform and a cost that never goes down. Your own store invests once and leaves recurring costs low. At Senda Lógica that store is Online Store, published at USD 2,500, with catalogue, search, filters, a page per product, and an admin panel with no product cap.

Fundamental decision 2. Payment methods. In Argentina, Mercado Pago is almost mandatory. Elsewhere in the region it is worth checking which method each country's customers prefer: in Brazil Pix dominates, and in Chile, Webpay. Offering bank transfer as an alternative captures customers who prefer to avoid card fees. In Argentina specifically it is worth considering multiple methods due to inflation and exchange rate fluctuation.

Fundamental decision 3. Logistics. In Argentina, Andreani, OCA, Correo Argentino, and Mercado Envíos are the most used options, with integrations available; it is worth comparing price and coverage for each area. Own logistics is worth evaluating if the geographic radius is limited. In Brazil: Correios plus private options. In Mexico: Estafeta, DHL. Shipping cost impacts conversion more than most people estimate.

Fundamental decision 4. Number of products. Fewer than 30 products: a simple store works. From 30 to 300: good internal search, categorization, and filters are needed. More than 300: requires serious architecture and possibly integration with a management system to keep stock synchronized.

Fundamental decision 5. Complementary sales channels. Marketplaces provide volume but reduce margins. Instagram and the WhatsApp catalogue work as a shop window that leads customers to the store or the chat, and TikTok Shop already operates in Mexico and Brazil. A multi-channel strategy usually beats a single store in revenue.

A typical ecommerce that works in LATAM: your own store as the centre, presence on one or two marketplaces, active social channels, consistent email marketing, and customer service over WhatsApp. The investment that matters in the first six months is usually not the store: it is sustained digital marketing.