How to Use Data to Decide What to Promote This Month

Every month, somebody at the restaurant has to decide what goes on the specials board, what the server mentions first, what gets a little extra attention on social media. Too often that decision comes down to whatever dish the owner personally feels like pushing that week, which is a perfectly human way to decide and also not a particularly reliable one.

The case for letting data pick, not preference

Personal preference is a weak proxy for what will actually resonate with your specific guest base, because the person deciding what to promote usually isn't a representative sample of one. An owner might personally love a particular dish for reasons that have nothing to do with whether typical guests feel the same way, memory, nostalgia, pride in a difficult technique. None of that is a bad reason to love a dish. It's just not a reliable signal for what to push this month if the goal is actually moving more orders and building genuine guest enthusiasm.

The dish worth promoting is the one already earning strong reactions from the guests who've tried it but hasn't yet been ordered by enough people to reflect that. That's a very specific, findable pattern in the data, and it has almost nothing to do with anyone's personal taste.

A simple filter for picking a monthly feature

• High like rate among guests who ordered it, ideally above your menu-wide average
• Order count that's noticeably below what the like rate would predict
• No recent negative pattern, like a spice or portion complaint, that a promotion push would just amplify
• Ingredients your kitchen can reliably scale if the promotion actually works

What promoting the right dish actually does

Promoting a dish that's already earning strong reactions creates a kind of compounding effect that promoting a mediocre dish never will. More guests try it, more guests love it, and the resulting word of mouth and repeat orders reinforce each other in a way that a forced promotion of a middling dish simply cannot replicate, no matter how much attention you give it. A promotion is an amplifier. It makes a good dish's momentum bigger and a mediocre dish's flaws more visible, which is exactly why picking the right one to feature matters as much as the promotion effort itself.

This is also a good discipline for avoiding promotion fatigue, where staff and guests both start tuning out specials because too many mediocre dishes have been pushed under the same banner over time. Reserving the spotlight for dishes the data actually supports keeps that spotlight meaningful.

What happens when the wrong dish gets promoted

Promoting a dish on instinct rather than data doesn't just waste the marketing effort, it can actively backfire. A middling dish pushed hard on social media or the specials board draws in more guests to try it than it would have otherwise, and a larger share of those guests come away unimpressed. That's a worse outcome than not promoting anything at all, since you've actively spent effort manufacturing more disappointed guests rather than more satisfied ones. The cost of picking wrong isn't neutral, it's negative.

Rotating what you promote instead of picking one favorite forever

Once you start using data to pick a monthly feature, it's worth resisting the temptation to just keep re-promoting whatever worked last time. The Promote list should refresh as order counts catch up to like rates, dishes that get discovered eventually graduate off the list and make room for the next underrated one. Treating it as a rotating queue rather than a single permanent pick keeps the spotlight moving to wherever it's currently needed most.

TasteGraph's Promote category exists specifically to answer this question: it surfaces dishes with a high like rate and comparatively low order count, the precise profile of something worth featuring this month, so the decision comes from actual guest reaction data rather than whoever's turn it is to pick the special.