Operator to operator.
No hype, no tool-of-the-week. Plain writing on custom software, modernization, data, and AI for the people who run growing companies.
Two reports.
All posts
The Guest Experience Gap: A 2026 LBE Field Study
Frogslayer deployed trained researchers to 50+ location-based entertainment venues as real guests. This is what breaks — and what it costs operators. Free PDF, no gate.
How Much Does Custom Software Cost in 2026?
Real ranges from a firm that publishes its prices: $25K–$75K for a utility, $150K–$750K for modernization, $750K+ for a custom ERP. What moves the number, and how to start for $3,500.
Signs You Have Outgrown Your Software
How to tell when QuickBooks, the spreadsheets, an off-the-shelf ERP, or a vendor's platform has stopped keeping up with your business, and what to do about each one.
How Long Does Custom Software Take to Build?
Realistic timelines for custom software by type of project, what actually drives the schedule, and why the first release should ship in months rather than years.
Software Project Rescue: The Warning Signs and the First 30 Days
How to tell a software project is in trouble before the deadline proves it, and what a competent rescue actually does in its first month.
Custom Software vs. Off-the-Shelf: A Decision Checklist
Twelve questions that tell you whether to buy a package, build custom software, or do the common third thing: buy the commodity and build the part that makes you different.
What Is a Software Audit, and When Do You Need One?
A plain explanation of a software audit: what gets examined, what you get back, when it is worth commissioning, and how it differs from a security scan or a code review.
Hiring In-House vs. a Software Firm: How to Decide
When a growing company should hire its own developers, when it should work with a firm, and the hybrid most companies actually end up with.
Offshore vs. Onshore Software Development: Which Is Right for You?
A straight comparison of offshore, nearshore, and onshore software development for growing companies: where each fits, what the lower rate actually buys, and how to judge the final cost.
Software Agency vs. Consultancy vs. Consulting Firm: What Are You Actually Hiring?
The labels blur together. Here is what each kind of software partner is built to do, where each one fails, and how to tell which one you are talking to.
Custom Software for Construction Companies: What Gets Built and What It Costs
The systems construction and specialty-contracting companies actually build: job costing, field data, estimating, scheduling, and reporting across the systems that never talk. With typical budgets and where to start.
Custom Software for Logistics Companies: What Gets Built and What It Costs
The systems freight, warehousing, and distribution operators actually build: operations visibility, dispatch and driver tools, customer portals, and the integration layer across a TMS, WMS, and accounting. With typical budgets and where to start.
Technical Debt in the AI Era
AI coding tools make it faster than ever to take on technical debt without noticing. What technical debt looks like now, why the bill comes due faster, and how to tell the healthy kind from the kind that stops a business.
How to Pay Down Technical Debt as You Go
A practical playbook for owners and executives: how to make technical debt visible, how much of each release to spend on it, what to fix first, and how to hold a team to it without becoming an engineer.
All Systems Fail: Is Your Company Outgrowing Its Systems?
What worked for you yesterday is failing you today. The symptoms show up in three places — management, employees, customers — and ten questions tell you whether it's time to build.
When Your Custom Software Stops Working for You
Recurring bugs, missed releases, a team that ships the wrong thing. What we find under the hood of struggling custom software, why it's usually salvageable, and what to do first.
Are You Ready to Start Your Custom Software Project?
Four things to settle before you build: the value, the risks, the budget, and who is actually committed. Plus what to do if you can't answer them yet.
What is Pre-Project Consulting?
How Frogslayer's pre-project consulting works: a Discovery Call, a Deep Dive with a ballpark range, a fixed-fee planning step, then a proposal. Why we do it in that order, and what you get at each step.
Questions to Ask a Software Development Partner Before You Sign
The questions that separate a firm that will finish your software from one that will bill you for trying. Ownership, team, method, communication, references.
Five Red Flags That Mean You Have a Team Problem
Bugs that come back, deadlines that slide, a partner who doesn't understand your business. Five signs the problem isn't your software but the team building it.
Horizontally Scaling Your Development Team
Why relying on one indispensable developer is a recipe for disaster, how to design a codebase that new people can join, and how to bring them on without slowing everyone down. Revived and updated from a two-part 2021 series.
The True Cost of Offshore Development
Offshore rates look like savings until you add onboarding, communication lag, rework, and what you lose when nobody on the team understands your business.
Build vs. Buy Software Analysis for Growing Companies
When off-the-shelf software is the right call, when custom is, and when the answer is neither. A plain-language build vs. buy analysis from a firm that builds.
Stalled AI project? We do rescues.
Roughly 80% of AI projects never reach production. The demo worked, then it died on a shelf. Here's how we rescue stalled projects and take over from a vendor who couldn't finish — without the blame game.
What to do before your software project starts
The decisions to make, the people to line up, and the data and access to gather before a custom software project kicks off. Plus what you can skip.
How to Validate a Product Idea Before You Build It
A practical method for validating a software product idea without writing code: the questions to answer, how to size the market, how to interview real users about the problem, and how to test the solution with mockups. Revived and consolidated from a four-part 2021 series.
Product vs. Platform: A Thing, or a Thing for Things?
Every vendor calls its product a platform. Here is what actually makes one, why it matters to your business model, and the questions to answer before you try to build one. Revived and updated from 2021.
Why We Build a First Release, Not an MVP
Minimum viable products have become checklist-driven, bug-laden, and undifferentiated. Here is why we aim for a first release instead, the three myths about it, and how to prioritize what goes in. Revived and consolidated from two 2021 posts.
The questions we'll ask you on the first call
What happens on a Discovery Call with Frogslayer: the questions we ask, why each one matters, and what it sounds like when a project is not a fit.
How we set a budget for a software project
Fixed budget, controlled scope, a strong first release. How we set the number for a software project, what moves it, and why we won't quote before design.
Judge the Final Invoice, Not the Hourly Rate
Why our hourly rates are higher than a freelancer's or an offshore team's, how we set them, and why the number that matters is what a working system cost you in the end. Revived from a 2018 outline.
How to Set a Custom Software Budget
The first question we get is how much a project will cost. The better question is how much it is worth. Here is how to set a budget you can hold to.
Fixed budget, fixed price, or hourly: who carries the risk
The three ways to contract for software, where the risk sits in each, when Frogslayer uses which, and why we recommend the model instead of asking you to pick.
Is Custom Software CapEx or OpEx?
How custom software lands on the books depends on what it is for, how long it lasts, and which phase of the work you are paying for. A plain-language guide.
Who Pays for Bugs in Custom Software?
Every custom system ships with bugs. Here is who fixes them, who pays, how it changes before and after launch, and how to budget so none of it is a surprise.
Budgeting for software maintenance
Why working software still needs a budget: what maintenance includes, how reserved hours and hosting are priced, and when maintenance becomes modernization.
Custom Software Maintenance, Explained
The four kinds of maintenance every custom system needs, where the line between upkeep and new development sits, and how to pay for it without surprises.
How to Write an RFP for Software Development
Most software RFPs select the best-looking document, not the best partner. What to ask instead, how to check references, and a simple fill-in template.
A Sample Software RFP Response, Walked Through
A hypothetical regional distributor's RFP, answered two ways. How a good partner handles budget, scope, and unknowns, and how a weak one does not.
The true cost of hiring your own developers
Salary is the small part. Recruiting, ramp, management, turnover, and the single point of failure. When an in-house team is right, and when a partner is.
Why Planning Phases Never Become Development
Most stalled software projects die between the plan and the build. The four reasons it happens and how a VDP Sprint is designed to get past them.
Waterfall, Agile, and How We Actually Work
What waterfall and agile were each trying to fix, why almost nobody practices either as written, and the parts worth keeping on a real project. Revived and updated from a two-part 2021 series.
Why Software Development Is a Team Sport
The myth of the lone programmer, who actually needs to be on a software team, why the team should run with autonomy, and what a cohesive team can do that a collection of contractors cannot. Revived and updated from 2021.
Managing Risk in a Custom Software Project
Every custom build carries risk, and the engagement model decides who holds it. How fixed price, T&M, and fixed budget with controlled scope compare.
What Happens After Development?
Nobody talks about the part after the first release: hosting, bugs, the next phase, and what it all costs over time. Here is how to think about the life of your software after launch, and how to budget for it. Revived from a 2018 outline.
Don't Sacrifice User Involvement for a Surprise Launch
Secrecy and user feedback are not mutually exclusive. How to keep the big reveal while making sure the software you reveal is something people actually want.
Velocity: Fast, Furious, and Flawed
Your dev team reports velocity every sprint. Why that number says almost nothing about whether the project is on track, and what to ask for instead.
Design Is Not an Up-Front Activity
The mockups are approved and handed to engineering. That is a milestone, not the finish line. Why design runs through the whole build and what it costs to stop.
Software Is Never Done: Don't Build It, Grow It
Custom software isn't built, delivered, and finished. It's grown. Why every system needs a plan for after launch, and what that plan should include.
What Is Technical Debt, and When Should You Take It On?
Technical debt trades long-term goals for short-term gain. How to spot it without reading code, when it is the right call, and how to keep it from stalling you.
You built a prototype, not a product. That's fine.
AI-generated and no-code MVPs hit a ceiling. What your prototype proved, what production needs, and the honest next step: a System Evaluation or a rescue.
Before an AI-Built System Runs Your Business
More companies are building their own systems with AI tools and one capable person. Here is what to check before that system becomes the backbone of the company.
The Surprising Reason Most Legal Tech Solutions Fail
Law firms are spending more on technology than ever, and the results are underwhelming. The reason is simple: the people choosing the software are not the people using it. Here is how to fix the selection process. Revived and updated from 2021.
Why 80% of AI projects fail — and what to do instead
Most AI projects fail for business reasons, not technical ones. Here's why, and the operator's way to be in the 20% that ships.
Stop buying AI tools. Start building capability.
Another seat license won't change your business. The operators pulling ahead are building capability — systems, data, and adoption — not collecting tools.
The bottleneck-first framework for prioritizing AI
Your best AI use cases are hiding in your bottlenecks, not your tech stack. A simple way for operators to pick what to build first.
The 6 workflows every owner-led service business should automate first
Across hundreds of projects, six workflows produce most of the real ROI in mid-market service businesses — ranked by payback period, with realistic costs and time-to-value.
AI agents vs. workflow automation: the difference most owners miss
Vendors call both 'AI,' but agents and workflow automation cost different amounts, break in different ways, and fit different problems. Here's how to tell which one you actually need.
Why your best AI use cases are hiding in your bottlenecks
Most AI opportunity assessments start with the tech stack and ask what AI could do. That's backwards. The use cases that move the business are in the operational bottlenecks you've been working around for years.
The four levels of AI integration in a mid-market service business
Not every workflow needs the same depth of AI. Here's how to match integration level to the work — and why getting it wrong costs you six figures.
Build, buy, or borrow your AI capability?
Most mid-market operators default to buying off-the-shelf AI tools — and most are wrong. A practical framework for choosing between building custom, buying a tool, or borrowing senior expertise on retainer.
Internal AI lead, consultant, or both?
You've decided AI matters — now you have to staff it. The honest tradeoffs between hiring an internal AI lead, using an outside partner, and running both, including when the right move is to do nothing yet.
When to use an AI partner vs. an internal hire
The $200K Director-of-AI hire vs. the $30-120K partner retainer looks like simple math — until you look closely. A practical decision frame for mid-market owners, including when hiring is the wrong move and when it's the right one.
An owner's 30/60/90-day plan for AI
You don't need a 12-month strategy to get started. You need 90 days of disciplined motion. Here's the exact plan we walk through with new AI Office clients on day one.
The three stages of AI maturity for a mid-market service business
Most AI maturity models are written for the Fortune 500. Here's the three-stage version that actually matches how a $20M service business moves through AI — and how to tell what stage you're really in.
The two-week AI audit you can run on your own business
You don't need a $50K consulting engagement to know where you stand with AI. You need about 10 hours over two weeks. Here's the audit — the questions, the artifacts, and what to do with the answers.
AI risk and governance for the mid-market owner who doesn't want to read 80 pages
You don't need an enterprise framework. You need six policies, three roles, and a few checkpoints. Here's a defensible AI governance posture scoped for a mid-market service business.
The KPI dashboard for AI in a mid-market service business
Most AI dashboards measure usage, adoption, and hours generated — none of which a CFO can defend. Here are the metrics that actually prove AI is working, and how to set them up without a six-month BI project.
How to hire and manage AI agents like employees
The companies getting good at agentic AI aren't the ones with the most sophisticated tech. They're the ones treating agents like new hires — defining the role, supervising the work, and retiring them when they don't perform.
The ROI math on AI: how an owner actually measures it
Most AI ROI claims are vendor noise. Here's the math an owner can defend to a CFO, a board, or a sponsor — and the discipline that separates real returns from theater.
The AI compounding curve: why waiting a year costs three
The cost of delaying AI another year isn't one year of forgone value. It's the gap that opens between you and competitors who started earlier — and the real math is roughly 10x what the naive version suggests.
A real AI budget for a $20M service business
Most AI budgets you'll see are theoretical or vendor-flavored. Here's the actual money math for a $20M B2B service business — the three components, the total, and the ROI that justifies it.
Quarter by quarter: what year one with an AI Office actually produces
A four-quarter, composite look at a real AI Office engagement — the workflows built, the cash spent, and the value captured. Year one is foundation-building; the ROI catches up and overshoots in years two and three.
Inside a $12M industrial services company's first 90 days with an AI Office
You've seen the AI Office page and the pricing. Here's the real 90-day arc — discovery to production — in the operator's own terms, with the numbers it produced.
The workflow that got an owner 12 hours a week back
An owner of a $25M industrial services firm thought he had no realistic path to working less. Twelve weeks later, AI was running the parts of his job he hated most. Here's exactly what we built.
A day in the life of an AI Office engagement
What you actually get with a monthly AI Office retainer — week by week, hour by hour. The honest version, drawn from real engagements.
What AI-native operations actually looks like in a $20M service business
The phrase gets thrown around. Here's the concrete version — what changes in the operating model, the org chart, and the day-to-day for a $22M commercial services firm 14 months in.
What working with Frogslayer actually feels like, month by month
An honest, month-by-month account of the AI Office engagement from the client's seat — the cadence, the time it takes, the first ROI conversation, and what happens when something goes wrong.
Patterns that repeat: what we learned from hundreds of projects
Six patterns separated the AI engagements that delivered ROI from the ones that stalled. The honest summary from an operator who watched them happen.
The hidden cost of AI pilots that never reach production
The cash you spent on the failed pilot is the smallest part of the cost. Here's what else gets destroyed when an AI pilot dies on a shelf — and why most companies underestimate the damage by 5x.
The real cost of doing AI the wrong way
The sticker price of a failed AI project is usually less than 25% of the total damage. Here's where the rest of the money — and the years — actually go.
Lessons from watching owner-led companies do AI wrong
Six failure patterns we watched repeat across owner-led AI engagements in 2026 — and the three disciplines that prevent most of them.
The AI readiness diagnostic: 15 questions to ask before you spend a dollar
Roughly 80% of AI projects fail — usually because the business wasn't ready, not because the technology didn't work. These 15 questions tell you which side of that line you're on.
The difference between using AI and operating with AI
Almost every mid-market company is using AI. Almost none are operating with it. The gap looks small until you're a year into it — here's what actually changes.
AI literacy at the leadership level: what senior leaders actually need to know
Your leadership team doesn't need to be technical. It needs the conceptual clarity to make good AI decisions. Here's the six-concept literacy stack that fits a mid-market leadership team.
AI transparency with your customers: what to say and what not to
Your customers are starting to ask how AI is involved in the work you do for them. Here's how to answer honestly, what to put in your contracts, and what it costs you when you get it wrong.
Why your IT person can't own AI (and why that's the owner's problem)
Handing AI to IT is the most common, most expensive mistake mid-market operators make — not because IT isn't capable, but because AI isn't an IT problem. Here's who should own it instead.
AI is the new electricity. Your operating model is the wiring.
Electricity didn't change the economy until factories were redesigned around it. AI is at the same inflection — and most mid-market companies are still wiring their old factory layout.
2026 is the year mid-market service businesses fall behind if they wait
AI won't change the middle market in 2026 — it will separate it. Here's what's changing in the next 18 months, and why operators who wait pay twice: first in opportunity cost, then in catch-up cost.
A letter from the CEO: what we've learned in 100+ mid-market AI engagements
The honest version, not the polished one. What actually separates the mid-market companies that get AI to work from the 80% that don't — written for the operators we work with.
How we run Frogslayer on our own AI operating system
If we're going to ask clients to invest in AI capability, we should be at least as far along as we're asking them to go. Here's the honest version of what AI looks like inside our own operations — what's working, what failed, and what it cost.
Why the Texas middle market is going to pay off
A founder's point of view on why the next decade of AI-enabled operating leverage will be won by Texas middle-market businesses — and why we built and stayed here to be in the room when they decide.
What I wish I'd known about AI before we started selling it
Twenty-four months into building the AI Office practice, and hundreds of projects into the firm, here are the lessons I'd hand my past self — what I got wrong about tools, operators, pricing, and where the real moat actually lives.
Why I won't take a client whose champion isn't identified
An AI engagement without an internal champion fails by month seven, every time. Here's why I turn down good money when no one's named the person who owns the work.