Post-M&A IT integration for fuel and convenience retail
When one operator buys another's estate, I move the technology from two disconnected worlds into one integrated platform: POS, forecourt, back office, networks, payments, vendor contracts. Site by site, without a site going dark.
Approach
Every integration I run follows the framework below, from the day the scope is still being discussed until the day operations can run BAU without me. The six stages a programme goes through are the ones typically planned for, but the handoffs between those stages deserve extra attention, because that's where most failures occur.
-
01
Scope discovery
Find out what was actually bought. The asset list, the forecourt and the contracts usually disagree.
More
Discovery doesn't mean only taking inventory. It requires a reconciliation. What you were sold in the data room and the estate that is trading day to day are not exactly the same: local workarounds, undocumented integrations, hardware that quietly passed end-of-support, expired contracts on "auto-renewal".
The first step is to review the data you already have before doing new surveys. On one programme, a suffix in a terminal's model number told us its connectivity type, answering in one column a question that would otherwise have meant visiting hundreds of sites. Always verify first: at this point you're sizing a programme against an inventory you can't yet trust.
-
Handoff: discovery to design
Most design blind spots are inherited from an incomplete discovery.
More
If you design the "to be" from the documents only, without verifying the reality underneath, that design will carry the blind spots from the discovery forward, surfacing when it's too expensive to go back to the drawing board.
This isn't only technical: a dealer or supply agreement with a hard expiry, notice windows or exit penalties can influence the whole rollout sequence. I've had a single contractual deadline force an entire cluster of sites into one month at a fixed cadence. If you miss that during discovery, the design neglects a constraint that you'll have to retrofit later, or pay the consequences if you miss it completely.
-
02
Target design & adaptation
Decide what the estate should look like after conversion, and adapt the standard to the market instead of forcing it.
More
Companies usually have a pre-thought design they want to implement after conversion. The hard part is not finding that out. The challenge is to decide how much to standardise and which step-outs to allow: if you force the standard overall, you break how a market actually trades, think of legal and fiscal compliance, fuel-card schemes, loyalty, payment rails, language. Customise too much and your operations team needs to support a zoo.
And not every site needs the same amount of change. On one programme I ran two scopes in parallel: sites needing a full platform migration, and sites already on target that needed only a product-identity change. Treating them as one scope would have overshot complexity. It's useful here to segment the estate by depth of change, not by geography or brand.
-
Handoff: design to planning
A carefully designed target with no clear path leading to it is just a drawing.
More
The to-be has to be calibrated against reality: the capacity you can secure across teams, the slot you can win in everyone else's backlog, the funding you can get, all inside the window the TSA covers you for. Skip that and you still have an elegant architecture and plan, but in practice it's unreachable before TSA end.
An important discipline here is to pre-align budget ownership before committing. Borrow budget from another programme without it, and the accounting problem will surely surface at the worst possible time.
-
03
TSA exit & conversion planning
Plan the exit from the seller's systems before the transition service clock runs out.
More
The TSA is a meter running against you. Every month on the seller's systems costs money and keeps you dependent on the party you just bought from. Try to rush the exit and you may hit the market half-finished. Planning is the negotiation that happens between goldplating and actually delivering on time.
This negotiation cannot be done successfully without a clear, documented agreement on what an MVP looks like, what will go into the post-MVP backlog, and who will manage that backlog later.
-
Handoff: planning to pilot
On a plan, a convenient pilot and a representative pilot look the same.
More
The plan assumes that the pilot represents the estate. That's the idea of a pilot, after all. Then the pilot gets chosen for what's easy to schedule, and the programme learns the wrong lessons, getting overconfident while playing on "easy" mode, which buys you a future problem.
Structure the pilot around your site archetypes, not around what's convenient. In a complex reality, engineering will have different priorities from the sales team. Those demands must inform the site archetypes, which will make the pilot stage more representative.
-
04
Pilot conversion
Convert a handful of representative sites first, where mistakes are cheap and the lessons are informative.
More
The first pilot site is important, but not worth over-focusing on. The site archetypes are the point. Structure the pilot in phases that cover the variety: unmanned, dealer-owned, special fuels, high-volume, complex forecourt.
And volume itself is a test: issues that stay invisible at one site surface at extended-pilot scale (concurrency, logistics, the edge cases that are more common than expected). The pilot's real job is not so much to succeed, but rather to find out what breaks, while breaking is still cheap.
-
Handoff: pilot to rollout
Agree on a definition of "good enough" before the pilot starts.
More
Deciding the exit criteria, i.e. when a pilot is good enough to scale, is a judgement call, hence it can't be made on the day. If you leave it until you're in the pilot, there's a risk you'll quietly move the line to wherever you're standing.
Scale with open issues and they scale too. A problem that's cheap to fix across three pilot sites is likely hundreds of times more expensive once deployment has progressed to hundreds of sites.
-
05
Estate rollout
Convert the rest at pace: reuse what worked, simplify to ensure repeatability.
More
By rollout, the limitation stops being technical and it becomes about people: getting a heterogeneous workforce to work the same systems in the same way. Under-train them and you'll drown in "How do I..." tickets on day one. In fuel retail, with high turnover, the people you trained during the pilot may be gone.
Every cutover is a live operation on a running business, often at night, repeated across crews, languages and regulations. Ensuring repeatability by making the steps known and understandable for everyone is key.
-
Handoff: rollout to handover
Converted is not the same as owned.
More
The most common way every programme ends badly: the team moves on to the next engagement before operations can keep the lights on. The knowledge disperses, and the ops team was never given the clarity, nor properly sized, to support what's been built.
It is a knowledge as well as a sizing question: how does the operation need to grow, in headcount and capability, for every site you've added to it? Answer that late and the estate ends up technically converted but operationally "nobody's problem".
-
06
Operational handover
Hand the estate to the people who will run it daily, on a condition, not on a date.
More
"Live" and "stable" are different states. The gap between them is hypercare and an agreed transition. When the trigger to demobilise is a date on the calendar rather than a condition that has been met, handover fails.
A clean exit needs a real definition of done: hypercare run properly, remaining problems resolved or handed over with named owners, a support model in place, ownership transferred to BAU. Demobilise on the condition being met, not a date on the calendar.
Six stages, five handoffs. The common pattern is that failure happens in the transitions between stages, and the fix is usually at the organisational level, not technical.
About

I'm Lisandro V. Robles. Independent consultant through LVR Advisory since 2025. Before that, six years at Shell running retail IT change across DACH and CEE: POS and payments, network migrations, managed-service transitions and post-acquisition integration, across ten European markets and 3,500+ sites.
Before Shell, I worked in investment banking operations at J.P. Morgan, and I bring that financial view into IT: every technology decision is also a business decision, and I weigh what it costs, what it protects and what it does to margin.
Today I lead integration and conversion programmes for fuel retailers in the Netherlands, Germany and Belgium, directly and in partnership with Haia Group. Remote-first, on site for the phases that need it.
English · Español · Deutsch · Nederlands
Start with a free consultation
Tell me where the programme stands and what's the biggest blocker keeping you awake. You'll get my take on it, whether or not we end up working together.