How to write an ecommerce replatform RFP

A commerce replatform is not bought the way most software is bought. Nobody stands up a storefront on a trial account and decides in a fortnight. The purchase carries a migration behind it that runs for quarters, so it clears whatever threshold a company sets for formal process, and it reaches an approver who is signing off a project rather than a licence. What follows is the document that survives that reading, assembled from the commerce interviews in our corpus.

How do you write an ecommerce replatform RFP?

When is a formal RFP required?

The trigger is a policy your company already has, not a judgment about the category. One industrial manufacturer describes a structured enterprise process requiring at least three proposals and a formal RFP for anything classed as an enterprise solution. A large consumer-goods manufacturer runs extensive requests for proposals and quotations because audit standards require it — the document exists to satisfy a control, and would exist regardless of what was being bought.

Others are looser, and the looseness is not carelessness. A subscription retailer describes an RFI or RFP handled collaboratively and often using spreadsheets, starting from leadership alignment that a third-party solution is needed at all. A regulated manufacturer sequences it the other way round: a personal assessment of the need, then IT, then a market evaluation, and an RFP only if that scan turns up viable alternatives. The scan comes first; the document is a consequence of finding something worth putting it to.

So the first question is not what to write. It is which of those you are in — because one produces a formal instrument with a response format and a minimum bidder count, and the other produces a comparison sheet.

Should you buy, compose, or build your ecommerce platform?

Buy tends to fit when time to market or internal development capacity is the constraint. Compose fits when requirements exceed what a managed platform covers. Build becomes a live option when vendor constraints are the problem and the engineering capacity exists to own the platform.

This is the section that has no equivalent in a CDP or loyalty RFP, and it belongs at the front. Across the commerce interviews, building or substantially assembling the platform is not a hypothetical — it is what several of the largest buyers say they are actively weighing, and in a few cases already doing. A B2B distributor frustrated by a monolithic incumbent names building its own platform as the primary consideration, driven by architecture flexibility and getting out from under vendor constraints. A specialty retailer running both consumer and business storefronts is deciding between deepening its existing investment, moving to a composable vendor, or an internally developed solution — three routes, one of which never reaches a vendor. A global manufacturer frames the whole exercise as capturing business requirements and then deciding build versus buy with enterprise architects, before any vendor conversation happens.

The reason to settle this before writing is that the requirements work is identical across all three routes and the procurement is not. Buyers who name a build option describe arriving at it through the same requirements exercise that would have produced an RFP — so the effort is not wasted if you end up buying. It is wasted if you write a vendor-facing document for a decision that was never going to reach a vendor.

What are the eight sections of an ecommerce replatform RFP?

Eight sections. The first and the last are where the interviews concentrate: buyers describe replatforms that chased a platform rather than a named outcome, and a contract term that later set their whole calendar.

What criteria matter most when choosing an ecommerce platform?

The three criteria buyers name most are total cost of ownership, integration with the existing backend, and scalability or architectural flexibility.

The grouping below is by how often buyers named each one as deciding their evaluation. These are frequency tiers rather than a strict ranking — the interviews support the grouping, not a precise order within it.

Grouped by how often buyers named each criterion as deciding their evaluation. Bars show the tier, not a weight or a vendor score.

How many vendors should you include in an ecommerce RFP?

There is no standard number in these interviews. The only explicit policy requires at least three proposals; other buyers describe shortlists of three to six, often narrowing to two finalists.

The corpus is thinner here than on criteria, and worth stating plainly rather than dressing up. One industrial manufacturer describes the only explicit rule: at least three proposals for an enterprise solution, then narrowing to the two best competitors on solution fit and price, with deep dives, demos and focus groups deciding between them. Others describe shortlists of three to six without naming a policy, and one runs a proof of concept against a named incumbent rather than a field.

What generalises is not the count but the shape: a long list assembled from what is already known and what the market scan surfaces, a narrowing to two on fit and price, and then a second, more expensive round of evidence-gathering on the finalists. Budget the second round properly. It is where the demos stop being presentations.

Who decides on an ecommerce replatform?

Two groups: a recommending group that evaluates the platforms, and an approving body that funds the migration. The split between them is consistent across the interviews.

The practical consequence is that one document has two readers. The recommending group is choosing a platform; the approving body is funding a programme. A requirements list that only satisfies the first can stall at the second, which is the failure mode buyers describe when they say executive management pushed the timeline back.

What role does procurement play in an ecommerce replatform?

In these interviews, procurement typically joins after the field has narrowed, handling contracting and commercial terms rather than choosing the platform. One buyer describes marketing and IT leading the evaluation before legal and procurement enter; another runs the technical evaluation and proof of concept first, then moves into procurement to settle contracts and commercial terms.

The consequence is a timeline one buyer describes with some feeling: contracting as lengthy and arduous, governed by company policy rather than by the deal. If your organisation has that reputation, the vendors will already know it, and the contracting window belongs in your plan as a real phase rather than a formality after the decision.

When should you start an ecommerce replatform RFP?

Work back from your incumbent agreement: in these interviews, contract expiry sets the replatform calendar more reliably than the business case does. Buyers describe having chosen a destination and waiting out a term to reach it, with target dates pinned to a renewal years out. One accessories brand extended its incumbent agreement specifically to create the migration window.

That inverts the usual planning question. The deadline is the renewal, the migration is the long pole, and the RFP is backward-planned from both. So the question that decides your start date is not how long the document takes — it is how long the migration needs, which is the number buyers most often get wrong in the optimistic direction.

What this RFP cannot cover

Two limits worth stating rather than papering over, because both sit where the risk actually is.

And a note on the instrument itself: no buyer here describes a weighted scoring matrix or a formal rubric issued to vendors. What they describe is criteria applied in prose, comparison held in shared spreadsheets, and a narrowing decided in conversation. The outline above is a requirements structure, not a scoring model. If you want the second, build it — but do not assume the category runs on one.

Common questions

Related reading

For the buyer-side lessons behind these requirements — why the migration rather than the licence is the budget line, and what the app ecosystem costs — see what buyers wish they'd known before replatforming. For what actually triggers the move, including the peak-season failure this page's closing section returns to, see why brands replatform. For the same document exercise in a different category, see how to write a CDP RFP.