Rebuilding a Small-Business Website with AI: What Must Survive the Redesign

Rebuilding a Small-Business Website with AI: What Must Survive the Redesign

GoDaddy's latest AI website-building update makes it easier to turn an existing site into a new design. For SMEs, the deeper challenge is preserving useful content, customer journeys and the business connections that a visual rebuild cannot automatically transfer.

A small business can outgrow its website long before it has the time to replace it. Services change, photographs become dated and a booking link points to a process that somebody has quietly stopped using. Yet the old site may still contain years of useful work: pages customers recognise, search traffic, product descriptions and connections to the systems that keep the business running. AI website builders promise to make renewal easier by turning a description into a working design. Their more consequential test is whether they help an established firm improve its online presence without losing the useful parts of what it already has.

GoDaddy’s 17 November product update for Airo AI Builder brings that problem into focus. It describes templates for starting a project and a rebuild flow that uses an existing website as a reference for a new version. The company explicitly distinguishes this process from automatic migration: catalogues, administration settings and integrations do not simply transfer with the appearance. For an SME, that qualification is more useful than a promise of an instant redesign. It identifies where AI can accelerate visible production and where the business still needs to plan, check and reconnect the operations behind the screen.

The first design is only one part of the job

In its 16 November announcement, GoDaddy described creating websites and web apps through written or spoken prompts. It also reported analysing more than 100,000 customer-created projects from a six-week beta period, with nearly one in ten classified as rebuilding or reimagining an existing site. These figures concern projects made inside the supplier’s service, not a representative sample of small businesses or evidence of commercial success. They nevertheless show why rebuilding deserves attention alongside starting from scratch. An established company has something to preserve as well as something to improve.

The attraction of conversational building is straightforward. Instead of learning every control in an unfamiliar editor, an owner can describe a desired change and inspect the result. That can reduce the distance between an idea and a draft. It may also make iteration less intimidating: a service page can be tried in a different structure, or a confusing description rewritten. But easier production does not resolve ambiguity about the business itself. If the owner has not decided which services are offered, how enquiries are handled or which prices are current, the builder has no dependable answer to recover. A polished design can conceal those unresolved decisions rather than settle them.

An illustrative independent repair firm might want a clearer website for customers seeking quotations. Its real problem could be that visitors do not know which photographs to send, whether collection is available or how long an assessment takes. A visual refresh alone would leave those questions unanswered. A better brief would specify the customer’s next step and the information needed to support it. AI could then help organise explanations and produce an initial layout around that purpose. This is a possible use to validate, not a reported customer outcome. It demonstrates why a business should describe the work a website must enable before describing how the website should look.

Rebuilding the surface, reconnecting the business

The distinction becomes concrete in GoDaddy’s rebuilding instructions. They say the existing content and branding guide a fresh design, while products, catalogues, store settings, payments and other integrations do not carry over. They describe reviewing and refining the generated version before publication. A business can therefore use the existing site as a starting reference without assuming the resulting project is a complete operational copy. This is particularly important for a small online shop, whose most valuable functions may sit behind an unremarkable product page.

A storefront is connected to decisions about stock, delivery, returns, customer records and payment handling. Recreating the page does not establish that these decisions have been recreated correctly. The same applies to a service business: a convincing appointment button might open the wrong calendar, omit a necessary question or reach an inbox nobody monitors. Before switching, somebody must follow the customer’s journey through to the actual business outcome. Can an enquiry be received and assigned? Does the relevant colleague see the information? Is a payment recorded where the firm expects it? These are questions about continuity of work, not cosmetic quality.

This does not mean every small firm needs a large migration project. The amount of work should reflect the complexity of the existing site. A sole trader with a few information pages and an external contact link faces a different task from a retailer with a catalogue and active orders. A sensible assessment would identify what the current site contains, what must remain available and which functions can be simplified. Some old features may no longer be useful. Others may look minor but be essential to regular customers. The aim is to make those choices deliberately, so that a generated replacement does not decide them accidentally.

Preserving the routes customers already use

A website has an address structure as well as a design. Customers may arrive through a saved link, an old email or a search result pointing directly to a particular service. When a rebuild changes those addresses, the business needs a way to lead visitors to the corresponding new page. Google’s guidance on site moves recommends mapping old URLs to new ones, testing the new site and implementing appropriate redirects. It also warns against sending many unrelated old pages to a single irrelevant destination. This is contextual guidance about changes in web addresses, not a claim that Airo performs these steps automatically.

For an SME, the implication is practical: the customer should not have to understand that a redesign has taken place. A link to a specific repair service should still help someone find that service. If the business keeps its domain but changes page paths, continuity still needs checking. If it changes hosting or domains as well, the scope becomes broader. The owner should establish which settings the chosen platform supports and when specialist assistance is worthwhile. AI may make producing the new pages easier while leaving this transition work intact. Ignoring that difference would confuse the speed of creating a draft with the readiness of a replacement.

The same reasoning applies to information presented within the pages. A generated rewrite may subtly strengthen a promise, omit a qualification or introduce a service the firm does not provide. Images can suggest facilities or products that are not actually available. Someone who understands the business needs to review these details, not merely approve the overall impression. A customer using a phone should also be able to read the content and complete the intended action without unnecessary difficulty. The useful standard is whether the new site helps people understand and deal with the real business, rather than whether it resembles a professionally styled example.

A smaller release can produce better evidence

The easiest way to control the scope is often to begin with a limited part of the site. A firm could prototype a revised service section or a campaign page before considering a complete replacement. In an illustrative test, the team might compare whether customers can identify the right service and submit the required information on the old and proposed pages. This would not establish an overall return on investment, but it could reveal confusing wording and missing steps early. Keeping the existing service available during review also gives the business room to correct problems without making customers part of an unfinished experiment.

The cost assessment should cover more than the advertised speed of the first build. AI credits, subscription conditions, catalogue preparation, reconnection work and staff review may all matter, depending on the chosen arrangement. The owner should verify the current plan and features rather than infer total cost from a free preview or a launch announcement. It is equally reasonable to compare a rebuild with repairing the existing site. A business whose main problem is one outdated page may gain little from moving everything. A business whose current system makes every change difficult may have a stronger case. The scale of the intervention should follow the problem.

Measurement then needs to connect the website to work the firm recognises. More visits are not necessarily more suitable enquiries, and more submitted forms may create additional handling without generating useful business. A small team could look at the completeness of incoming requests, the number requiring clarification and whether customers successfully reach the next step. Seasonal demand and other marketing activity can influence the result, so improvements should not automatically be credited to AI. The purpose of measurement is to make the decision less speculative. It should reveal both where the new site helps and where the process still creates effort for customers or staff.

The useful promise of an easier rebuild

The wider importance of conversational website building lies in changing who can participate in improving a business’s digital presence. An owner who can describe a problem may be able to explore a workable draft sooner, while a designer or developer can focus on the parts that require expertise. That is a potential change in the division of work, not proof that professional help has become unnecessary. The more the site touches transactions, records or complex integrations, the more valuable it becomes to know where the generated draft ends and the operational system begins.

GoDaddy’s November update is therefore best read as an opportunity to reduce one barrier to renewal. It does not demonstrate that an existing business can be moved intact through a conversation, nor that a new design will automatically attract better customers. For a small firm, the strongest use of AI is to make improvement easier while keeping responsibility visible: decide what should change, preserve what still works, test the connections and release when the customer journey is ready. A website is useful because it helps people do business. Faster production matters when it serves that purpose without breaking the relationships and processes already built around it.

Image: Lee Campbell, Designer’s two-screen setup, via Wikimedia Commons; CC0 1.0 Public Domain Dedication. Illustrative photograph, unrelated to GoDaddy or a specific customer; resized for web display.

Related Posts