← All guides
8 min readBy Mike Petroskey

Is it you, or your tech stack? What we covered in our 2027 tech webinar

Every week somebody asks Justin or me some version of "I'm using this, should I switch?" So we did a whole webinar on it. The recording is below. If you'd rather read, here's the gist. The two AI prompts I promised are at the bottom.

My thesis

I told everybody up front that the point was to prove my point, and if they disagreed, to say so. Either I persuade you or you convince me I'm wrong.

Here's the argument. The feeling that there's a crazy tech race going on isn't new. It's where we were always headed. Your technology isn't necessarily holding you back, either. It's set up to give you exactly the results you have right now. If you want a different result, you start from the result and work backward through your tools, your process and your people. And you don't necessarily need anything new to do better. You might. You just might not.

Who we are

Justin and I run However We Got Here. You can call us fractional CTOs or sales and operations people, but what we really are is a channel partner for the specialty trades. In the old days, if you wanted a server, there was a Dell guy and an HP guy, and their advice leaned toward whatever they sold. We work with all of it, so we can tell you what's good about a tool and what isn't. Sometimes the answer is a tool we don't partner with. Sometimes it's the one you already have.

StashGrade started as the diagnostic we used with our own clients, and it turned into something owners can run on their own.

How we got here

The robot helped me with this part. I wrote out eight moments in history and asked it what I was missing.

Writing started as accounting: bushels of wheat and who was working on the pyramids that day. Barbed wire is when fencing became an industry, because now you could fence in thousands of acres without building a wood fence around all of it. Then the telephone, then computers that took up a whole office floor, then the cloud, then the iPhone. Around 2011 to 2015 you get Jobber, JobNimbus, ArcSite and CompanyCam, each one going after a piece of the workflow nobody else was covering. COVID pushed everybody to e-signatures and online payments. Then ChatGPT.

Under all of it is computing power doubling every 12 to 18 months while the price keeps dropping. When I started my career I was selling 78 gig drives for about $20,000 apiece. Now there's a terabyte in your phone and nobody thinks twice. AI does the same thing one level up. Building software used to be for coders. Now it's for anybody with an idea.

Four kinds of CRM

We used CRMs as the example because everybody has feelings about theirs.

Enterprise: Salesforce, Microsoft Dynamics, SAP. Gold standard. Nobody buying your company is going to knock the price down because you run Salesforce. But you're looking at six figures to set it up and somebody on payroll who knows how to run it. If you're at $30 million and growing, look at it. If not, keep going.

Contractor suites: Jobber, JobNimbus, ServiceTitan. Built for the trades, not for fencing. Can they work for a fence company? Absolutely. Quotes, crew calendars and invoicing are all in there, and for a lot of companies they cover most of what you need most of the time.

Configurable: GoHighLevel, Zoho, Monday, Attio. Built to fit any workflow. Cheaper to start and super flexible, but the add-ons add up, and field work takes some setting up.

Fence specific and DIY: built by fencing people for fencing people. Usually it's an owner who didn't like anything out there and built their own. Then their friends wanted it too.

Here's my prediction. You'll see a bunch of new fence operating systems at Fence Tech this year, and about double that next year. Building it got easy. Running a software company didn't. What happens when it breaks at 7 a.m. on a Monday? Who handles a security problem? When we decided to sell StashGrade, we handed the software side to a professional software company, because with financial data we didn't want that risk sitting with us.

And think about selling your company someday. A buyer wants a business that makes them money from day one. If the whole thing runs on homegrown software and one guy who knows how it works (go call my nephew Don), expect to hear about it in the price.

Is it the software or your people?

Somebody asked this one live: "Everyone on my team blames our CRM, but I'm not convinced that's the problem."

I hear it all the time. Either the system can't do it, or it can and your people aren't using it that way, or it's some of both. Figure out which before you buy anything. Switching tools is easy now from a technology standpoint, and it's hard from a business standpoint. You don't get unlimited tries with the same crew. Find something that gets you most of the way and keep refining it.

What we're seeing in our benchmark data: the companies still growing in 2026 keep finding problems, trying fixes, checking whether the fix worked and trying again.

What to do about it

Test the tools you already pay for. Beat them up before you replace them, and be direct with your vendors (nicely).

Stay connected. Keep talking to other owners, because that's where the good ideas come from. Play with AI, but don't let it lead you down a path. Tell it what you think you want to do, then ask it to beat you up.

Execute and refine. Don't go on a bunch of DIY adventures and wear out everybody in the office. If you can't measure it, don't bother, because how would you know it worked? And you will fail some of the time. AI is a lot better than it was last year, but it's not perfect. Don't give up.

The homework: map your workflow with AI

Map how a job moves through your business, from the first call to the Google review. Copy Prompt 1 into ChatGPT, Claude or Gemini. It interviews you one step at a time, then builds a one-page map of your tools, how information gets from one step to the next, and what to fix first. Give it 20 to 30 minutes, and sit down with your office manager. They know which steps get typed in twice.

Prompt 1: Build the map

You are helping me map how a job moves through my contracting business, from the first lead to the review. I want a one-page document I can check every month. Interview me first. Ask about ONE step at a time, wait for my answer, then move on. Keep questions short and plain. If I don't know a number, write "don't know" and keep going. The nine steps: 1. Lead comes in 2. First reply 3. Appointment 4. Measure and quote 5. Sign and deposit 6. Schedule and install 7. Invoice and get paid 8. Books and job cost 9. Review and repeat. For each step, ask me: Which tools do we use here? (software, spreadsheets, paper, text messages, "in someone's head" all count) Who does it? Roughly how many in the last 90 days? What information has to get to the NEXT step? How does it get there: by itself, typed in by hand, or not at all? When the interview is done, build the document with these parts. A. MY WORKFLOW: the nine steps in order. Under each: the tools, the 90-day count, and every hand-off to the next step marked as MOVING (goes by itself), BY HAND (someone retypes it), BROKEN (does not get there, or nobody is sure) or NO TOOL (nothing covers this step). B. SCOREBOARD: one line at the top: how many hand-offs are moving, by hand, unsure and broken. C. FIX FIRST: the three problems that cost me the most, ranked. For each: what is going wrong, a rough guess at what it costs me in hours or dollars per month (label it a guess and show the math), and the simplest fix. Prefer fixes that use a tool I already pay for over buying a new one. D. NUMBERS TO CHECK MONTHLY: five or six numbers I can pull myself, where to find each one, and what "good" looks like. Include at least: share of jobs with a known lead source, leads to quotes, quotes to signed jobs, share of signed jobs that match an invoice, and days from install to paid. E. QUESTIONS I COULDN'T ANSWER: everything I said "don't know" to, with who in my company to ask. Rules: Plain words. No jargon. Write for a busy owner reading on a phone. Do not recommend new software unless a step has no tool at all. Do not invent numbers. If I didn't give one, say it's missing. Format it so I can paste it into a Google Doc. Start by asking what trade I'm in and how many jobs we do in a typical month. Then begin with step 1.

Prompt 2: The monthly check

A month later, open the same chat (or paste last month's map into a new one) and paste this:

Here is my workflow map from last month. Ask me for this month's numbers from section D, one at a time. Then tell me: what got better, what got worse, which hand-off I should fix next, and whether any "by hand" step has grown enough that it's worth connecting the tools. Update the document and show me the new version.

Pay the most attention to section E, the questions you couldn't answer. That's where nobody in the company is watching.

Want the easy button? StashGrade builds the map from your real numbers instead of your best guess, straight from QuickBooks and your job software. It's free to start.

Next month

Justin's running the next one: is your ad money working, and how do you fix it? Most owners we ask can't say what they get back for what they spend, so we'll show you how to start tracking it. Date coming soon. If you've got questions about this one, or something you want us to cover, reply to any of our emails. Good, bad or otherwise, I want to hear it.

See how your KPIs stack up against a healthy business

Connect QuickBooks and get your margin by job type in minutes. Free to start, read-only.

Connect QuickBooks for free

More for your trade: Fencing, Decking, Roofing, Concrete.