What makes an advisor contest pay for itself

We've run a lot of advisor contests with our dealers: key service challenges across an entire dealer group, tire and alignment spiffs at single stores, and month-long penetration contests on filters, wipers and spark plugs. Most of them made money. A few didn't. And a few looked like they did, right up until we checked the numbers properly.

Here's what we've learned about building one that pays for itself, and how to prove that it did.

Size the opportunity before you pick the prize

Before deciding what to pay, work out what's actually on the table. Take the store's current penetration on the service you want to push, and compare it to where the store was last year and where comparable stores are. Multiply that gap by the store's customer pay RO count and the gross on each service. That's the gross available, and it tells you how much of it you can afford to hand back to advisors.

Then put it in terms an advisor cares about. "Alignment penetration is down" doesn't move anyone. "If we just get back to last year's alignment rate, there's this much bonus available per advisor" gets attention.

Pick one behaviour, and pay on it the way advisors think

The contests that worked focused on one service, or a short list of related ones. The payouts that worked were simple and tied to each sale:

• A set amount per unit, like a flat payout for every customer pay RO with tires sold.
• A kicker for the attachment you really want, like an extra amount when that tire RO also includes an alignment.
• A higher tier that unlocks at a penetration target, so your top advisors don't coast once they're comfortably in front.

Show counts as well as percentages. Managers almost always ask for the count, and "four more alignments" means more to an advisor than "two points of penetration".

Pick metrics where the team is close together. If one advisor has the contest locked up by the fifth of the month, everyone else stops trying. When a manager asks us what to run next, we usually look for a metric where the advisors are neck and neck.

Measure against what would have happened anyway

This is where most contest results go wrong. Comparing the contest month to last month makes almost every contest look like a win, because stores have seasons and trends.

We project what the store would have done on the contest services if its pre-contest trend had simply carried on, year-over-year growth included. The contest only gets credit for the gross above that line, and only on the services it targeted. If a store was already trending up before the contest, it would have kept growing without one, and that growth isn't the contest's.

Contest lift = actual gross on the contest services − projected gross on those services, based on the pre-contest trend

Contest profit = contest lift − payouts

Run this mid-contest as well, not just at the end. A check-in halfway through gives leadership time to act on what it shows.

Separate the contest from the store

In group contests, sometimes one store comes in well under its projection. When we've dug in, it usually wasn't the contest. The shop had filled up with warranty and internal work, or more lines were being deferred or left pending, so there were fewer customer pay hours for the advisors to sell into.

Pull those stores out of the contest math and give them their own conversation. Otherwise one store's bad month hides the fact that the contest worked everywhere else. It can also go the other way, where one strong store makes a weak contest look good.

Make it visible every day

The contests with the best engagement had standings that updated daily on each advisor's own page in KPI Sight. Single-store spiffs worked well in weekly periods ending on a Sunday, while group challenges ran for the whole month. The things that kept people engaged:

• Managers pushing their own team up the leaderboard.
• Stores competing against each other, not just advisors within a store.
• Wildcard draws, which give advisors outside the top three a reason to keep selling.

Make sure every advisor can actually see the contest on day one. An advisor who can't find the page isn't in the contest.

Check the data before anyone gets paid

Most contest disputes we've dealt with turned out to be real data problems, and the advisor raising it was usually right. The usual culprits:

• Pay type. Counting warranty and internal ROs inflates penetration. Lock the contest to customer pay work.
• Op code mapping. A service that isn't mapped makes a store look like it stopped selling it. The same op code typed in different cases can be missed, and an op code mapped to the wrong service can inflate a count.
• Unsold lines. Quotes and declined lines that carry a contest op code but were never actually sold.
• The date range. On the first of the month, a month-to-date view shows nothing, and it looks like the contest broke.

Before a payout, open the top advisor's ROs and confirm the services were sold on customer pay work.

Rotate it

The same spiff, month after month, stops moving anything, and the manager usually knows it before the numbers do. Change the focus with the season, like tires and alignments heading into tire season, or point it at the store's weakest area right now, like one-line ROs or a maintenance item that's dropped off.

Then watch the month after the contest ends. If the rate holds, you built a habit. If it drops straight back to where it was, you paid for a month of effort, and the next step is coaching, not another contest.

How KPI Sight does it

We build the contest on each advisor's page from the same repair order data as the store's dashboards, so the number an advisor is chasing is the number the manager sees. We handle the setup, rollout and leaderboard, and when it wraps we report the lift against projection store by store, so you know whether it paid for itself.