The short version
- The workshop’s reputation was strong offline and invisible online. That gap was the whole project.
- We rebuilt the site from scratch rather than patching, and gave the team a CMS so they never have to call us for an edit.
- Local search was the priority: schema, metadata, and content aimed at people typing “car repair in Migori”.
- Every decision here applies to any service business outside Nairobi.
Most case studies you read are victory laps with a graph at the end. This one is a description of decisions, because the build is recent and we are not going to invent traffic numbers to make a point. What we can tell you is what we changed and the reasoning behind each choice, which is the part that transfers to your business anyway.
On the word “rebuilt”. We did not patch an existing site. We rebuilt the business’s digital presence from nothing, designing and building the whole thing from scratch. The live result is at smartlakemotors.com, and the full project summary is on our work page.
The brief
Smartlake Motors offers servicing, diagnostics, genuine spare parts, and roadside assistance as a certified Isuzu centre. In the workshop, none of that is in doubt: the certification is real, the parts are genuine, the mechanics know the vehicles.
Online, a driver with a failing Isuzu somewhere between Migori and Isebania had no way to establish any of it. The gap was not quality; it was legibility. The brief was to build something that matched the professionalism of the workshop, made a stranger trust it in about eight seconds, and made booking a service easier than making a phone call to a number they did not have.
1. A design built for this business, not adapted from a template
We designed and built the whole front end from scratch (home, about, services, team, gallery, and contact) mobile-first, because virtually every visitor arrives on a phone, often on mobile data with patchy signal.
Templates fail this kind of business specifically. An automotive template assumes a dealership: stock listings, finance calculators, model comparisons. A service centre needs entirely different things forward: what is wrong with my vehicle, can you fix it, where are you, can I come today. Structuring the site around those questions rather than around a template’s existing pages was the single highest-leverage decision in the project.
The gallery earned its place for the same reason. Photographs of the actual workshop, the actual bays, and the actual team do more for trust than any amount of copy, and they are the one thing a competitor cannot copy.
2. A dashboard the team runs themselves
We built a CMS dashboard so the Smartlake team can update services, publish posts, add gallery images, and read enquiries without touching code or contacting us.
This is the decision most businesses under-value at quoting time and most appreciate within a month. Without it, every small change (a new service, a price update, a photograph from last week’s job) is an email to a developer, a wait, and a small invoice. Sites without a CMS go stale within a year, and a stale site is a trust problem as well as a search-ranking one.
If updating the site requires us, the site will not get updated. That is not a criticism of the client; it is how every business works.
3. A blog aimed at the questions drivers actually type
We built “From The Workshop”, an SEO-focused blog system for vehicle-care articles.
The strategic reasoning: a service centre can only rank for so many service-page keywords, and those are the most contested terms available. But the questions car owners ask before they book (how often to service, what a warning light means, whether a part is genuine) are asked constantly and answered badly. Publishing real answers, written by people who repair these vehicles daily, earns visits from drivers who have a problem right now.
It also solves the freshness problem. A workshop that publishes a couple of useful articles a month is demonstrably active, to search engines and to a human deciding whether the business still exists.
4. Booking that lands where the team already looks
Booking and enquiry flows hand off to email, so a request reaches an inbox the team already checks and can reply to quickly.
This is deliberately unglamorous. A full booking engine with calendar slots and confirmations is a genuine web application and would have been the wrong first spend for this business. Service scheduling here depends on bay availability, parts, and the specific vehicle, all things a person establishes in one exchange faster than any form can capture.
So the site’s job is to start that exchange with as little friction as possible, and to make sure it does not get lost. Cheaper to build, faster to answer, and it fits how the workshop actually runs. It can grow into a scheduling system later, once there is volume to justify it.
5. Built to be found in Migori, not in general
A complete SEO and schema overhaul: metadata, structured data, and on-page SEO built for local discovery, so a driver searching for car repair in Migori finds Smartlake rather than a national directory listing.
Local search is where the entire commercial case for this site lives. A business on a highway serves people who are nearby and have a problem now. That means:
- Structured data describing the business, its location, and its services as machine-readable facts.
- Page titles and descriptions written for the searches that matter, not stuffed with the town name.
- Location, road, and service area stated naturally in the content, the way a person would say them.
- A page per service, so a specific search lands on a specific answer.
None of this is exotic. Almost none of it is done by the competition, which is precisely why it works.
6. Analytics configured before launch, not after
GA4 was set up from day one, so the business can see what is working and where its next customers come from.
Analytics added six months later can tell you what is happening now. It cannot tell you whether the new site changed anything, because there is nothing to compare against. Configuring it before launch costs almost nothing and is the difference between managing the site and guessing at it.
What transfers to your business
Strip out the automotive specifics and five decisions remain that apply to any service business outside Nairobi:
- Structure the site around your customer’s questions, not around the pages a template happens to include.
- Buy the ability to update it yourself. It is the difference between a live site and a brochure that ages.
- Publish the answers you already give on the phone. Your expertise is your cheapest marketing channel and you are giving it away unrecorded.
- Match the contact method to how you actually work. The most sophisticated option is often the wrong one.
- Win local search before national search. It is less contested, and it is where your customers are.
If you want the underlying framework rather than the example, we wrote it up separately as the five-step chain a lead has to survive. And if you are working out what this kind of build costs, start with website costs in Kenya.
Common questions
Why not use a template for a service business website?
Because templates are built around a different business. An automotive template assumes a dealership, with stock listings, finance calculators and model comparisons. A service centre needs to answer what is wrong with my vehicle, can you fix it, where are you, and can I come today. Structuring the site around those questions rather than around the pages a template happens to include was the single highest-leverage decision in this project.
Does a small business website really need a CMS?
In almost every case, yes. Without one, every small change, whether a new service, a price update or a photograph from last week’s job, is an email to a developer, a wait and a small invoice, so the change does not get made. Sites without a CMS go stale within a year, and a stale site is a trust problem as well as a search-ranking one.
Is a blog worth it for a garage or workshop?
Yes, for two reasons. A service centre can only rank for so many service-page keywords, and those are the most contested terms available, whereas the questions customers ask before they book are asked constantly and answered badly. It also solves the freshness problem: a workshop that publishes a couple of useful articles a month is demonstrably active, both to search engines and to a human deciding whether the business still exists.
Do I need a full online booking system?
Usually not as a first spend. A booking engine with calendar slots and confirmations is a genuine web application. At Smartlake, scheduling depends on bay availability, parts and the specific vehicle, all things a person establishes in one exchange faster than any form can capture, so booking and enquiry flows hand off to email and reach an inbox the team already checks. It can grow into a scheduling system later, once there is volume to justify it.
How do I get my business found in local search?
Win local search before national search, because it is less contested and it is where your customers are. That means structured data describing the business, its location and its services as machine-readable facts; page titles and descriptions written for the searches that matter rather than stuffed with the town name; the location, road and service area stated naturally in the content; and a page per service, so a specific search lands on a specific answer.
Run a service business that deserves better online? Tell us about it. We are open to new work, and the next case study on our work page could be yours.
Wilfred Gweth
Wilfred co-founded ByteFrame in 2025 and handles design and front-end engineering on every build, including Smartlake Motors. Meet the team.