One of the most destructive attitudes in the modern workplace is the “yeah, yeah yeah.”  Have you heard it?  The research guys put some numbers together, and marketing puts some numbers together, and manufacturing put some numbers together.  Everyone comes together, preaches their numbers, and says, “yeah, yeah, yeah” to the other guys in the room.  I find this behavior as commonplace as I find it disrespectful and passive aggressive.

Net result, each participant leaves the meeting with the intention of using his/her own numbers, and management lets it happen in an effort to pick other battles.  Particularly in the case of forecasts: we know the forecasts are a shot in the dark anyway, right?

The first crack in the dam

This attitude–in my opinion–is the first crack in the dam.  When marketing doesn’t make their numbers, the first objection is that the numbers weren’t agreed in the first place, or that manufacturing couldn’t ramp up fast enough or that HR lagged on hiring.  Or worse (and I have seen this), marketing DOES make their numbers, but manufacturing didn’t buy them in the first place, didn’t ramp up their operations, and they are stuck defending—sometimes effectively—that they didn’t buy someone else’s numbers.  Ask them if they challenged those numbers and insisted on alignment.

You see, lack of unity in numbers and definitions render the debate dead on arrival.  What you need is be able to look at your planned versus actual in a sober, dispassionate way and not provide excuses for responsible parties to wiggle out.

A unified forecast

The good news is that BI has gotten dead easy, provided you take what programmers call an “object-oriented” approach.  It is the simple-sounding but terribly-difficult task of defining EXACTLY what you mean by terms thrown around so casually, such as “utilization” or “yield.”  First you define these objects with no equivocations in your BI landscape, then you insist that most of your metrics are calculated.  If yield is “food weight ready to cook / food weight purchased” then you should only need those two measures and not a third in your system.  You would be surprised how often a metric that should be calculated is loaded as a separate bit of information, and how rarely the resulting numbers tie out.

Collaborate, then lock it down

There should be no tolerance for dissent once a forecast is agreed.  That is because if your forecast is constantly shifting, you have nothing to measure against.  If you’re saying that you’re more agile then that, I would argue that the forecast period for your decision-making should be coordinated with the lead times for those decisions (another post to follow).  To wit, if you need a 6-month window to significantly change the size of departments, don’t adjust forecasts every two weeks, right?

By locking the forecast down, you encourage your team to debate the forecast and really come to an agreement during that process.  It isn’t always civil, but once you take the time to unify that forecast, you can manage with everyone on the same page, and next period you get smarter.  Isn’t that the point?