Codehesion
Let’s talk
Case study · Tiger Wheel & Tyre
RetailLogistics & automotiveProject rescueCode auditLegacy modernisation

New features on top. A rewrite running underneath.

We took over the code and launched their member app, and it has stayed live through every release since.

Let's talk →Tell us about a project like this.
    03

    A launch date, and an app that had to be ready

    Tiger Wheel & Tyre's member app is how customers join their membership, manage their vehicles and claim their benefits. When their previous partner let them down, they called us in with a launch already committed and members using the app every day.

    04

    Too live to stop, and too important to get wrong.

    There was a launch commitment to keep, and members were using the app every day.

    1. 01

      Reliability Members subscribe, pay and claim through the app. A release that broke something broke it in front of them.

    2. 02

      Speed to change The business had more it wanted to offer members, and every change had to land without risk.

    3. 03

      Ownership The business wanted every part of the product in its own hands.

    05

    The impact we made

    The app was too live to stop, so we didn't stop it.

    A rebuild would have meant close to a year with nothing to show, so we ran two tracks in the same sprint.

    New features went on top, against the business's own priorities. Underneath, every service a feature touched was rewritten to our standards as it went. The rewrite was paid for out of the same capacity as the features, so the roadmap never waited for the clean-up, and the clean-up never waited for a quiet month.

    06

    Every feature left the code safer to change.

    Here is how the work ran, and why.

    Two weeks, three codebases, one report

    Before we advised anything, three of our engineers spent two weeks inside all three codebases: the mobile app members use and the two services behind it. The result was a written report. It told the business what it had, what it could keep, and why a full rebuild was the wrong answer for an app with a launch commitment and an active member base. Advice came after the audit, never before it.

    Features and the rewrite, one backlog

    Features were delivered against a backlog the business prioritised, and the rewrite followed them: when a feature touched a service, that service was rewritten in the same sprint. That one rule keeps the rewrite moving without ever needing a budget of its own.

    The safety net before the speed

    We didn't accelerate until we could see what we were doing. Before speeding up, we put the safety net in place: automated tests, monitoring, safeguards on the services and one code standard for every change. Feature flagging let new work go live switched off and be turned on when the business was ready. Speed came after that, and it held.

    One domain at a time

    Each rewritten domain follows the same clean architecture. The point isn't neatness. When every domain has the same shape, an engineer who knows one knows them all, and each domain we finish is safer to change than the one before.

    Less to protect

    We kept the payment flow as lean as it can be: the less a system holds, the less there is to protect. Sign-in and access control were reviewed in the same pass.

    Plumbing they own

    Messaging, file transfer and vehicle lookup run on services the business owns, and every service is documented, so the product and the knowledge of how it works belong to the business.

    Attached to their team

    We cover business analysis, UX/UI design, development, QA and release management. The core pod is two developers, a project manager and a QA engineer, scaled up when the business needs more. We meet their technical stakeholder every week, demo the work, test in their stores and release through their standing change-advisory process. They decide what matters next. We get it live.

    07

    A small change is now an ordinary ticket.

    More than 1,300 tickets later, a request from the business goes into the backlog and goes live. Here is what members can do now.

    1. 01

      Join A free entry tier sits below a three-tier membership, so a customer can start without paying and step up when it suits them.

    2. 02

      Pay Traffic fines, licence disc renewals and vehicle cover, all paid in the app.

    3. 03

      Claim and redeem Car wash cash back, and loyalty vouchers that can be partly redeemed and the rest kept for next time.

    4. 04

      Stay close A store finder with maps, push and in-app notifications, biometric sign-in, vehicle-sale enquiries that reach the business as leads, and a line into the call centre.

    And the whole app has a new look, delivered on the same running product.

    08

    Two more products followed.

    Alongside the member app, the business asked us for two more products: an app for their branch teams, and a platform for their corporate customers.

    The corporate platform was built clean from our own templates. On our own internal code-quality rating it scores 95%.

      New features on top, the rewrite underneath, and you stay live the whole time.

      09Let’s talk

      Yours is live too. You don't have to switch it off to fix it.

      Tell us about a project like this.