"Where are we with Harbourline?" "Good. They're keen. Should land this month." "Great. Next one." Forty seconds, and nobody in the room learned anything. Do that eight times and you have held a pipeline review. The deal that is landing this month has been landing this month since June, and everyone at the table knows it.
That is not a seller lying to you. It is a meeting built to produce exactly that answer. If the only thing a review measures is whether a deal is still alive, the cheapest way through it is to keep every deal alive.
A status report is a performance
Most pipeline reviews ask one question in eight different costumes: is it still on? Everyone in the room understands the scoring. Saying "it's moving" costs nothing. Saying "I think this one is dead" costs you the deal, a chunk of your number, and twenty minutes of being asked what you are going to do about it.
So the pipeline stays fat and the forecast stays wrong, and neither of those is a discipline problem. You built the scoreboard. People are playing to it.
Half the room is not describing the same thing
Gartner surveyed 243 chief sales officers and senior sales leaders between November and December 2024. Forty-nine percent said their sales organisation's definition of a qualified lead differs greatly from marketing's. In the same survey, only 45 percent reported that their organisation had met several of its 2024 strategic goals.
Those are big-company numbers, and most New Zealand tech companies look nothing like the businesses in that sample. Read it as the mild version of the problem rather than the severe one. If an organisation with a sales operations function cannot agree on what a qualified lead is, the six-person team where the founder, the one seller and the technical co-founder each carry a private definition is not in better shape.
Two people can look at the same deal and honestly report different things. Until your stage names describe something the buyer did, rather than something the seller feels, the review is a conversation about vocabulary.
So what do you actually ask?
Stop asking about the deal. Ask about the buyer.
"What did they do last?" Not what they said, what they did. Took a meeting, sent the contract to their lawyer, asked for a reference, went quiet for eleven days. Actions are checkable, and they are the one part of a deal the seller does not control.
"Who else has to say yes, and have we met them?" Most stalls turn out to be a person rather than a reason. This question finds them while there is still time to go and meet them.
"What would have to be true for this to close by the date on the board?" This is the one that does the work. It asks for a chain of events, out loud, with names and dates in it. A seller who believes the date will give you the chain. A seller who does not will hear themselves saying it and move the date without you having to ask.
Then cover the thing the meeting never covers: what would have to be true for this not to close. A review that never produces a slipped date or a closed-lost is not a review. It is a roll call.
The CRM will not do this for you
Salesforce's State of Sales report for 2026, an anonymous survey of 4,050 sales professionals across 22 countries with New Zealand among them, fielded in August and September 2025, found that sellers spend 40 percent of their time actually selling. The other 60 percent goes on non-selling work, manual data entry among it.
So your pipeline data is written by people who are already behind, in a system they know is read mainly to check up on them. Tidying that record does not make it true, it makes the fiction neater. The review is where the record gets corrected out loud, by the person who knows, which is why the questions you ask decide what the data is worth.
In a market this small, the founder is usually in the room
That changes the meeting. In a five-person company the person hearing the update also set the price, wrote the product and signed the lease. Bad news is not a status change, it is a judgement on someone sitting two metres away, and it will be delivered accordingly.
Two things help, and I use both. Move the review off the deals and onto the buyers, because "walk me through what Harbourline actually did last" is a much smaller thing to answer honestly than "is it going to close". And make killing a deal something that scores. Close one out early in the meeting, thank the person who called it, and let the room watch it cost them nothing. That buys you one honest review. Do it four weeks running and you have an honest pipeline.
So look at your own board this week. What is still on it purely because nobody has been given a safe way to take it off?
Is your pipeline telling you the truth?
In about 6 minutes, the free Growth Scorecard shows where your sales are leaking and what to fix first.
Get your free scorecardSources. Qualified-lead definitions and goal attainment: Gartner, "Gartner Survey Reveals Less Than Half of CSOs Report Their Organization Met Several 2024 Strategic Goals," press release, 21 May 2025, reporting a survey of 243 CSOs and senior sales leaders conducted from November through December 2024, in which 49 percent said their sales organisation's definition of a qualified lead differs greatly from marketing's and 45 percent reported that their organisation met several of its 2024 strategic goals. Note the base: that sample is chief sales officers and senior sales leaders, so it describes organisations considerably larger than a typical New Zealand tech company, and it is used here as the mild case rather than the severe one. Seller time split: Salesforce, "State of Sales" report for 2026, an anonymous survey of 4,050 sales professionals across 22 countries including New Zealand, fielded August to September 2025 using third-party panellists, reporting that sellers spend 40 percent of their time selling and the remaining 60 percent on non-selling work including manual data entry. One widely quoted figure was considered and cut. Gartner is often cited for the finding that fewer than half of sales leaders and sellers have high confidence in their organisation's forecasting accuracy, which would have opened this piece neatly, but that number comes from a February 2020 press release that names the survey without publishing a sample size, a respondent breakdown or fielding dates, so it is not used here. The three questions, the rule about scoring a killed deal, and the observations about founder-run reviews in New Zealand come from my own work with New Zealand tech founders rather than from published data.