From 5 Hours to 30 Minutes: Fixing Your Estimating Process
If you're spending four or five hours on an estimate and only one in five converts, you're paying about $750 a week for the privilege of chasing work.
An estimating database fixes that. It's the least glamorous system I build, and it's the one that gives contractors their weekends back.
I've built estimating databases for every area of construction except civil, plus a handful of service industries. My actual skill isn't estimating. It's seeing patterns and fixing the flow.
The problem every service contractor has
Talk to service contractors, the ones on the service side of construction, and nearly all of them struggle with the same thing: getting estimates out on time.
The reason is simple. There are so many variables that it's hard to break the work into steps you can reuse.
I loooove hearing, "Little lady, it's not that easy to build something to make this simpler."
Look, sir. I haven't spent the past three decades doing exactly that so you can tell me I don't know what I'm talking about.
Is it complicated? Yes.
Is it impossible? No.
Hot tub trouble
You get the call. Somebody went to the State Fair and bought a hot tub. Now they need it hooked up. Simple thing. They've done the heavy lifting. You just have to connect it.
You and I both know it's not that simple, so you schedule a visit.
You talk to the owner, take photos and measurements. Then you look at the meter. It's an old Federal Pacific, and between the safety concerns and the fact that you can't get parts for the new breaker, you're now looking at a meter swap on top of the circuit. Also, the hot tub is a lot farther from the panel than he said on the phone.
You break the news. He's disappointed. He mentions his nephew is an electrician and could get him a great deal on a disconnect.
You're happy for him. You promise the estimate by end of week.
Back at the office, it's worse than it looked on site. You're sketching it out and the material list keeps growing. Five hours later you've got an estimate that covers everything except material procurement and drive time, because those always get missed on complicated service calls.
It gets accepted two weeks later, and you feel pretty good.
You shouldn't. That estimate cost you most of a workday, and four out of five like it won't come back at all.
An estimate is just a bucket of money
When you've been building estimates as long as I have, you realize that's all it is. Your job is making sure the bucket is big enough to cover the work so you're not paying the customer for the privilege of working for them, while staying competitive enough to win.
That's what an estimating database does.
It's exactly what it sounds like: a collection of material dollars and labor hours attached to tasks. You pick the task, add it to the estimate, and you've got a number.
Imagine handing that homeowner a price while you were still standing in his backyard. You'd probably have walked away with an approval, even though his nephew could've saved him twenty bucks on a disconnect.
Pre-built assemblies are the whole trick
You build "general" activities with average materials and labor.
You know roughly how much LT you need. You know you need an outdoor-rated disconnect and receptacle, and a connection to the tub, which is a separate service call once the tub is actually installed.
Set your standard parameters:
What are typical runs in your area?
Do you usually trench, or do you have a lot of covered decks you can run under?
Then build your assemblies. For this job, three of them:
Standard hot tub electrical hookup
Conduit and wire, by the foot
Standard 100A meter swap
That's your estimate. Add a standard term saying that anything you couldn't see gets charged at your billable service rate, and that you'll tell them before you proceed.
Because sometimes swapping one of those old Federal Pacific meters is simple, and sometimes it's a $@#%show.
"But I can't order material off it"
Here's what drives field guys nuts. They want to dump the estimate straight into a material order, and with assemblies you can't.
Sounds like a dumb system, right?
Wrong. Because not every estimate converts.
Customers shop. They're gathering numbers to build a budget. They're deciding whether to hire it out or try it themselves (and then usually make a bigger, more expensive mess, but they tried).
For every five estimates, maybe one becomes a job.
So why are you doing a full material take-off five times to win one?
The math
Say each estimate takes about four hours: drive there, take photos, talk to the customer, drive back, download pictures, compile notes, price it, write it up. It takes longer if you let it sit four weeks before you sit down with it.
Say you're doing five a week, because you're small but growing.
That's 20 hours a week. At $50 an hour, that's $1,000 a week spent chasing work. No fuel. No software. Just your time.
Now flip it.
With assemblies, estimates take an hour. I call it an hour to be conservative, because once your assemblies are dialed in, you'll knock a lot of them out in 30 minutes on site.
Five estimates at one hour is $250 a week.
When one converts, then you do the full take-off for the material order. Call that two hours, or $100.
New total: $350 a week instead of $1,000. You just saved $650 a week, and you can spend it chasing more work or not working on Saturday.
Software you can actually use
I've built estimates in Accubid, ConEst, ServiceTitan, Pricebook, Jobber, Housecall Pro, Service Fusion, a few I've forgotten, and Excel.
Yes, Excel. I've built many, many estimating workbooks in Excel, and for a specialty service or a small shop it works fine.
Most of the real platforms are expensive with serious monthly fees, and there's a reason for that. Without good estimates converting into jobs, your business doesn't exist.
You can only win so much work on a handshake and a good attitude. Eventually even the Good Ol' Boy code stops carrying you, unless you're the only provider in town. In which case, carry on. You're set.
How to start
Look at your own numbers first. You already have past jobs to learn from.
Are you tracking materials by job? (QuickBooks does this. Sort of.)
Are you tracking hours by job?
How's your truck stock?
How's your warehouse stock?
How many material trips do you typically make per job?
Here's the part nobody expects. Building the database makes everything else more efficient too. You start planning more. You start tracking more. And you end up with growth that's repeatable instead of accidental.
Start with your five most common jobs. Build an assembly for each. That alone will cover most of what comes through the door.
FAQs
What is an estimating database? A stored set of tasks with material costs and labor hours attached. You select the tasks a job needs and the estimate builds itself, instead of pricing every job from scratch.
How long should a service estimate take? Once you have assemblies built, often under an hour, and frequently a few minutes on site. Without them, four to five hours is common for anything complicated.
Can I build an estimating database in Excel? Yes. For a small or specialty shop it's often the right starting point. Dedicated platforms are worth it as volume grows, but the discipline matters more than the software.
What if the job turns out to be more complicated than the assembly? Include a standard term stating that unforeseen conditions are billed at your service rate, and that you'll notify the customer before proceeding. That's what protects the assembly approach.
Frankie Allen spent 30 years in construction, from drafter to journeyman electrician to project manager on residential, commercial, hospital, and federal jobs. Rippling Roots provides fractional assistant project management to contractors in Alaska, Hawaii, and the Western U.S.
