Confession post. Some background so you know where this is coming from: I co-founded a VC-backed logistics marketplace, we raised 8 figures over a few rounds, and I ran the commercial org as COO. At peak that was 30+ sellers and account managers across two markets, all running through one CRM that I was personally accountable for at board level. So when I say the pipeline data was bad, I mean I was the one presenting that bad data to investors every quarter.

A few years in I got so fed up with stale deals that I did the thing every advisor tells you to do. Tied money to it. Small bonus component for data completeness, fields filled, activities logged, stages current. Reps keeping clean data finally get rewarded, messy ones shape up. Made total sense in the spreadsheet.

Within a month the CRM was 100% complete and maybe 40% true.

Fields stuffed with placeholder junk. Activities bulk-logged Sunday nights from memory, you could literally see the timestamps clustered between 9 and 11pm. One rep copy pasted the same call notes onto six different deals. I was paying people to make the data LOOK maintained. Killed it after a quarter, and undoing the damage took way longer than that because now I couldn't tell which historical data was real either. Try building a board forecast on that.

The part that stung: my best closer had the worst data. And he was right. At 7pm after a day of calls, typing notes into a database is genuinely the lowest value use of his time. The guy was correctly prioritizing. The bonus didn't change that math, it just changed what the corner cutting looked like.

What actually helped, ranked by impact:

Cutting required fields from ~15 down to 5. Boring, free, biggest single win. Half our fields existed because someone wanted a report once in 2019 and nobody ever deleted them.

Weekly deal-by-deal reviews, me or a team lead. This worked. It also ate 5-6 hours of senior time every week and collapsed the month I was traveling for a fundraise. Renting a human sync engine, basically.

The bonus. Negative impact lol.

The pattern I never escaped: everything that worked needed a senior person burning hours enforcing it. Everything that tried to automate the motivation (KPIs, dashboards, the "CRM is our source of truth" speech I gave at least four times) did nothing or backfired.

I've compared notes with a lot of founders in the 10-50 seller range since and it seems near universal. Big enough that bad data actually costs you real money in forecasting, too small to justify a RevOps salary.

So genuinely asking: anyone here solved pipeline hygiene at that size without it hanging on one person's weekly discipline? And has anyone else tried paying for data quality directly? Did it backfire the same way or did we just design ours badly?