EN Every customization you agree to costs you nothing today and nothing this month — the bill lands on delivery, support and upgrades, and on whoever inherits the account two years from now. That's why reps systematically over-customize: it isn't a character flaw, it's a structural one. The upside is immediate and yours; the cost is delayed and someone else's, and people discount delayed costs steeply even when they do land on them. So before you nod, ask four questions: what outcome is this change actually for (people describe an implementation when they mean a result); can a standard workaround get eighty percent of it; are you willing to pay separately and wait separately for it (price is the cleanest filter there is); and has one customer asked or have ten. One customer means customization. Ten means your roadmap has a hole.
EN Adding feels risk-free: one more option, one more bullet, one more slide — worst case it's just unused. Wrong. Extra information isn't neutral. Impressions behave more like an average than a sum, so a hard piece of evidence standing next to three mediocre ones doesn't read as four points, it reads as one diluted point. And the more options are on the table, the more likely the winner is "not yet." Cut proposals to one recommendation (two at most, and they must differ obviously on one axis — expensive-and-fast versus cheap-and-slow, not three near-identical tiers). Run the deletion test on every claim: if removing it wouldn't change the decision, it isn't a claim, it's noise, and noise charges rent. Then dare to recommend — "we can do all of these, but given what you just told me I'd only recommend this one, because…" shows range and converges the decision in the same sentence.
EN A sentence carries information only to the extent that some people couldn't say it. When every vendor's homepage promises an end-to-end integrated solution that empowers growth, the phrase's signal strength is zero — buyers filtered it into noise years ago. Two kinds of sentence still get through. First, results concrete enough to be falsified: "a retailer your size went from a five-day close to a day and a half" survives where "highly efficient" does not. Second, an explicit exclusion — "we don't do cross-border settlement, you'll need someone else for that" — which works precisely because it is expensive to say and people who are bluffing won't say it. The test takes one minute: read your own pitch aloud and strike every sentence your biggest competitor could lift verbatim. Whatever is left is actually yours.
EN Diagnosis has a price: it costs the buyer's time, stretches the cycle, and quietly signals "buying from these people is a project." When someone already knows what they want, the requirement is standard, and the amount is small enough for one person to sign, every extra round of questions loses you points. Match effort to the complexity of the decision, not to your professional pride. Three signals to shut up and quote: they open with price, lead time or contract terms; they can state the exact spec, model or seat count; and the spend is small enough that nobody else has to be convinced. Two or more, and you compress the whole discovery into a single question — "the only thing that might change the recommendation: do you need multiple locations connected?" — then send the number. The asymmetry matters: failing to diagnose costs you one deal, while diagnosing when nobody asked costs you your reputation as someone who is easy to buy from.