In 2017 I bid a parish roadway project in Louisiana for $2.1 million. The next guy was at $2.8 million.

Thirty-three percent on the table.

Here's how it happened. We needed work. I let our lack of backlog cloud my judgment. I didn't pull our costs from similar jobs we'd already built, and I didn't look at a single bid tab from a comparable project. Two things that would have saved us a lot of money and pain.

When the bid results went out to the whole company with my name on them, my heart sunk and I wanted to crawl under my desk.

Then it got worse. I was the PM as well as the estimator, so I got to go build the job I'd just bid. We got clobbered. We wrote it down six figures.

So: left a fortune on the table, and lost money on the work I bought with it. Expensive lesson. Worth every dollar, eventually.

I was new to asphalt estimating, and I got better at it. But that job was the start of me actually paying attention to my competitors.

It wasn't only me

I'd love to tell you the problem was one green estimator having a bad day.

The truth is we hated losing more than we liked winning. That was the culture, and culture shows up in the numbers. It showed up in ours, and where it showed up worst was our percentage left on the table.

When you hate losing that much, every bid gets a little shave you can't quite justify. Nobody decides to give away margin. You just find yourself on the low side of every call, job after job, and at the end of the year you wonder where the money went.

FMI trademarked a phrase for this: “Volume Kills, Profit Thrills”.

The pressure behind it is real, though. There's a balance in the asphalt business. The fixed costs of a production facility and laydown resources require volume and tons to get to break even, full stop. You're balancing the need for backlog so you don't run out of work against trying to maximize what the market has to offer. It's a complex puzzle.

But needing tons is not the same as needing those tons at that price. I didn't lose that money because our plant needed volume. I lost it because I never checked what the market was willing to pay.

What we started doing

There is nothing that sophisticated about this.

We grabbed the tabs from the last year, or the year before. Found the jobs that look like the one we were bidding. Ran those prices against this bid's quantities. Adjusted for today's conditions. Ran the averages. Adjust based on gut and feel.

One job at a time. One competitor at a time. One line at a time.

It took me hours. And I got pretty dang close.

Eventually our whole team was doing it for every job and it turned into a fun guessing game each letting.

My grandfather was doing a version of this before anybody called it “data analysis”. He didn't have a spreadsheet. He had the tabs, the quantities, and enough sense to know that the guy who beat him last month was going to price asphalt about the same way this month.

People are predictable

That's the real lesson in the tabs. Companies bid the way they bid. They load mobilization or they don't. They're always high on structures and always sharp on grading. They get aggressive in Q1 and careful in September.

Those habits show up in the numbers, bid after bid, and once you see the pattern you can forecast it.

Part of an estimator's job is estimating the project correctly. The other part is predicting what the other guy is going to do. Most people only do the first half.

Until somebody scrambles the signal

Your competitors are readable right up until they aren't. Some of them know you're reading the tabs, and they'll scramble the signal on purpose - move money a different way, move where they hide their margin, throw a couple of bids way off so the averages stop meaning anything.

Often it isn't even on purpose. A company hires a new chief estimator and the whole personality of their bidding changes. They open a plant twenty miles closer and their haul structure is a different animal. They get burned on a structures job and go gun shy for two years. Or they get thin on backlog and start reaching, which is exactly what I did in 2017.

When that happens, the history you have on them is no longer data. It's noise.

So use your judgement. When a competitor breaks their pattern, you can’t average across the break. Draw a line, start a new baseline, and remove the anomalies. Gut and feel still matters greatly.

Your cost is your cost

None of this means you bid off somebody else's number.

Your cost is what it takes you to do the work. Your price is what the market will let you charge. Bid tabs tell you about the market. They tell you nothing about your cost - and if you let a competitor's unit price talk you off your own number, you deserve what happens next.

I know, because the $2.1 million version of that mistake was mine.

The questions I ask on every job

• What does history say about where this bid lands?
• What do I think each competitor will price it at?
• How accurate have I been against these guys before?
• Has anything changed about them that I need to account for?
• What does the data say, what does my gut say, and when they disagree, why?
• What happens if I'm wrong?

Write your picks down before bid day. Compare them to the tabs afterward. That feedback loop is the whole thing. It works on one and only jobs too.

"We don't bid off tabs. We bid our costs plus a fair profit."

I've heard that plenty. Those two things aren't in conflict.

You can build an accurate cost, take the margin you deserve, and still use history to sanity-check where you're landing. Not everything is predictable. People do crazy stuff on bid day. Markets change fast. You'll lose a close one now and then.

But if you treat every bid as a hypothesis and check your work afterward, you get better. That's not a theory, that's just how getting good at anything works.

The job that made me a believer

We had a competitor we used to outbid with ease. For years.

Then they got acquired by a bigger, more sophisticated outfit that understood exactly what I'm describing here. Within a few months their whole approach changed.

That's a scrambled signal, and I didn't adjust for it. I was still bidding against the company they used to be.

They got us by $7,000 on a $5 million job eight miles from our plant.

I was not happy. And then I got to work. They forced us to sharpen our costs, tighten our takeoffs, and start forecasting their moves instead of reacting to them. They made us better.

Try it on five jobs

You don’t have to take my word for it. Go test it.

For five upcoming bids that matter to you:

1. Predict what each competitor will bid.
2. Ask what would have happened if you had priced just below the low number.
3. Then check the tabs. How close were you?
4. What did you nail? What you miss, and what does it tell you?
5. Run the delta across all five. What does that do to your bottom line for the year?

In an industry where the money left on the table hovers around 9%, there's always room for improvement.

The tabs are sitting there after every letting. You can read them, or you can keep wondering where the money went.

‍