Decision framework · vendor-neutral

How to choose a CRM platform: three questions, in order.

The tools differ less in features than in assumptions about your company. Answer three questions honestly and the shortlist writes itself — this page is the framework I use before any stack decision.

The platform is the residue of three answers

Platform selection usually starts with a vendor shortlist and a demo calendar. That is backwards. The tools differ less in features than in assumptions about your company — and buying a tool whose assumptions don’t match is how companies end up paying enterprise prices for systems nobody can run.

Three questions decide it, in order.

One — where does your truth live?

If the data that defines your audiences sits in a warehouse — joined tables, financial history, consent from several systems — you need a platform built for relational weight (the SFMC class). If your truth is what users do in the product — events, sessions, feature usage — you need an event-stream platform (the Braze class: Braze, Customer.io, Iterable).

Companies that answer «both» should read that answer honestly: it usually means the warehouse is the truth and the events are the trigger. Weight wins; see the SFMC vs Braze comparison for how that plays out.

Two — who operates it daily?

A platform is an operating commitment, not a purchase. The SFMC class needs someone who writes SQL without flinching and treats templates as code. The Braze class needs marketers who think in events and a product team that keeps the schema honest.

The test is one sentence: name the person who will answer for deliverability a year from now. If the name doesn’t come, the gap isn’t the platform — it’s the owner, and no vendor fixes that.

Three — which failure can you afford?

Every platform fails somewhere. The enterprise class fails slowly and expensively: unowned automations multiply, defects hide under polite dashboards, and the bill arrives years later. The product class fails fast and visibly: you outgrow the segmentation, schemas drift, and the ceiling announces itself mid-quarter.

Slow failure suits companies with governance and audit muscle. Fast failure suits companies that iterate. Pick the one your culture will actually notice — an unnoticed failure is the expensive kind.

Questions this page gets asked

How do I choose a CRM / lifecycle platform?
Answer three questions in order: where your truth lives (warehouse vs product events), who will operate the system daily (data-literate CRM team vs ticket-free marketers), and which failure mode you can afford. The platform is the residue of those answers — not the starting point.
Should I choose the platform before hiring the team?
No — it's the most common expensive mistake. The platform your team can operate beats the one with the better demo. Decide the operating model first: who owns segmentation, who writes the queries, who answers for deliverability. Then buy the tool that model can run.
How much does a wrong platform choice cost?
The license is the small part. The real bill: a year of half-built journeys, a migration project (untangling templates, data contracts, suppression logic), and the opportunity cost of programs not shipped. Getting the choice right is cheaper than any discount.
Do I even need an enterprise CRM platform?
Not always. If your lifecycle program is under ten journeys and your segmentation fits in the product database, a mid-class tool (or even transactional email plus a scheduler) buys you a year of learning what you actually need — at a tenth of the cost.
Who should make the platform decision?
Whoever will answer for its revenue a year later. If that person doesn't exist yet, that's the real gap — a platform without an owner produces campaigns without results, on any vendor.

Want the answer for your stack, not the general case?

The Diagnostic reads your data, your team and your programs — and gives the platform answer in writing, with the reasoning attached. Two weeks, fixed price.

Start with the €2,900 Diagnostic or read SFMC vs Braze first