ScenariosWorking the deal
The seat is taken. That is not the same as settled.
They signed with someone else, probably before you ever spoke to them. Most teams mark the account dormant and move on, which is exactly why incumbents keep accounts they no longer deserve. Every incumbent has a renewal date, a set of unmet promises, and at least one person inside who argued against them.
01What is really happening
Displacement is a timing problem disguised as a product problem.
The account is not comparing vendors today, so a feature argument lands on nobody. What changes the answer is a moment: a failed rollout, a price increase, a reorganisation, the departure of the person who chose the incumbent, or a new initiative the incumbent cannot serve. Miss the moment and you are early, which looks identical to being wrong.
- A better product does not dislodge a working one on its own.
- The person who chose the incumbent cannot easily admit a mistake.
- Being early and being ignored feel the same from the outside.
02What Adrata reads
The buying history, and where the incumbent is thin.
Adrata reads what the account has actually bought, who drove those decisions, and which of them have since left. It watches for the changes that reopen a settled question and holds the competitive picture for the specific incumbent rather than a generic battlecard.
- What they bought, when, and who owned the decision.
- Whether that owner is still there.
- The change that makes reconsidering reasonable rather than disloyal.
03The move
Be the obvious call at the moment the question reopens.
The recommendation is a route and a timing, not a takedown. The aim is to be already known and already trusted by the people who will be in the room when the incumbent comes up for review, which means the work starts long before the renewal date.
- The people to know before the question is asked.
- The angle that does not require anyone to admit a bad decision.
- The trigger to wait for, watched for you.
The model behind it
You should be able to argue with the recommendation.
A move you cannot interrogate is a move you will ignore the first time it is wrong. These are the models doing the work in this scenario.
Causal uplift, not correlation
Did the action cause the outcome, or did the good deals just get more attention?
Treatment effects are estimated with propensity weighting and doubly-robust estimators, adjusting for deal size, buyer-group size, whether the deal was intro-sourced, whether it was an existing customer, and known competitive pressure. Estimates that have not been adjusted for confounders are quarantined so they can never be presented as a cause.
Thompson Sampling on what to send
Which version of this actually earns replies, for this kind of buyer?
Message decisions (template, tone, call to action, send time, subject style, opening hook) are treated as arms with Beta priors and sampled rather than fixed. Where a workspace has thin data for a specific context, the estimate falls back through archetype and industry to a global prior instead of pretending to know.
Break-in probability
Of every possible way into this account, which one actually opens?
Each candidate route is scored on seven components: relationship strength, how relevant the connector is to the target, freshness, consent safety, social cost, evidence quality, and expected lift. A calibrated model converts them into the probability that the route breaks in, with each component’s contribution visible.
Where it runs
The surfaces this scenario uses.
Who this is for