Win/Loss Analysis for Competitive Replacement Deals

A replacement deal — where the buyer rips out an incumbent vendor to install you — is not a normal win. The buyer has already paid for the category, trained their team, integrated the workflows, and explained the choice to their boss. When they switch, something specific broke. When they don't switch, something specific held. The win/loss interview that treats both outcomes like a generic deal review will miss the only signal that matters: the trigger and the friction.

2.3x

Most win/loss programs were designed for greenfield. The questions assume the buyer was choosing between vendors for the first time. They ask about feature comparisons, pricing perception, and sales experience. Useful, but they miss the question that defines a replacement: what changed about the incumbent that opened the door? Without that, the program produces tidy reports that don't tell the CMO where to push.

What makes replacement deals different

Three dynamics separate replacement from greenfield, and each one demands a different question set.

The trigger isn't a need — it's a fracture. Greenfield buyers have a problem they're solving for the first time. Replacement buyers had a working solution that stopped working. The fracture is usually quieter than a feature gap: a renewal price hike, a champion leaving, a roadmap broken promise, a support ticket that took six weeks. If your interview opens with "what were you looking for in a vendor," you've already missed it.

Switching cost is the real competitor. In greenfield, the alternative is "do nothing" or "build it ourselves." In replacement, the alternative is "stay with what's working badly." The buyer is comparing your value to the incumbent plus the migration tax — data export, retraining, integration rebuilds, the political cost of admitting the original purchase was a mistake. Wins happen when the fracture cost exceeds the migration cost. Losses happen when it doesn't.

The reference point is asymmetric. A greenfield buyer evaluates you on a feature checklist. A replacement buyer evaluates you on whether you fix the specific thing that broke — and they discount everything else. Your AI roadmap doesn't matter if they're switching because the incumbent's reporting was slow. Your reporting needs to be visibly better, demonstrably, in the first demo.

The interview structure that surfaces the trigger

The standard win/loss script has three problems on replacement deals: it asks about your sales process before the buyer's incumbent relationship, it treats price as a single variable rather than a renewal-cycle pressure point, and it lets the buyer narrate the decision in retrospect rather than reconstruct the moment of fracture.

Here's the structure that works. Sixty minutes, recorded, with explicit consent. Don't run these yourself if you're the CMO — use a third party or an analyst the buyer doesn't recognize from the deal. Bias is real and it's worse on replacement.

How to code the responses

The interview is the data collection. Coding is where most programs fall apart — they tag every response as "pricing" or "features" and lose the texture. Replacement deals need a tighter taxonomy.

The right-hand column is what you act on. "Lost on price" tells you nothing. "Renewal-cycle price shock" tells you to time your outreach to the buyer's renewal date and lead with TCO over a three-year window.

What the pattern looks like in aggregate

After 30–40 replacement interviews, the picture sharpens. You'll see clusters — not random scatter. Most replacement programs end up with three or four dominant trigger types, and the distribution tells you what to build, what to message, and where to hunt.

The chart above is a composite — your distribution will differ. What matters is that you can produce one. If your win/loss program can't tell you the count of deals won on each trigger type, it isn't giving the CMO the input they need to allocate budget against the demand-gen, product, and battle card workstreams that respond to each pattern.

Acting on the signal

A win/loss program that produces a quarterly PDF and dies there is a cost center. The signal has to route somewhere.

Where each replacement signal goes

Check it off

0 of 0 done

PrintResetSaved in your browser

Prefer a printable version?Download PDF →

What to skip

Three things that look useful in a replacement win/loss program and aren't.

Don't ask the buyer to score you on a 1–10 scale across ten attributes. They'll give you tidy numbers that average to nothing. The scores are post-hoc rationalization and they wash out the actual texture of the decision. Replace numerical scoring with the counterfactual question.

Don't interview only closed-won deals. Closed-lost replacement deals — where the buyer evaluated you and stayed with the incumbent — contain the highest-value signal in the program. They tell you exactly what migration cost or feature gap your prospects can't yet stomach. Aim for a 60/40 won/lost split in your interview cadence.

Don't let sales run the interviews. They have an account-team relationship to protect, the buyer will hedge, and the bias is unrecoverable. Use a neutral interviewer. The cost is real but the signal is worth it.

We ran win/loss for two years before we separated the replacement deals into their own track. Once we did, the trigger map showed up in 40 interviews. Our pipeline built itself from that map.

The template

The interview guide, the coding taxonomy, and the action-routing matrix above all live in one Word document, designed to be filled out per interview and rolled up across a quarter. It's the artifact a PMM hands to a third-party interviewer and gets back with the data structured for analysis.

Run ten interviews against the template before you draw any conclusions. The first three will feel ragged. By the seventh, the trigger patterns will start repeating. By the tenth, you'll know which two or three signals deserve a quarter of your demand-gen calendar — and which battle card needs rewriting before Friday.