Knowing Which Competitor You Are Actually Competing With
Why this matters
Most owners can name fifteen firms in their trade within thirty miles and cannot tell you which three actually take work off the table. That gap costs money in both directions. It sends a shop chasing a rival that has never once appeared in a customer's decision, and it leaves the rival that appears in every second decision unnamed and unanswered. A competitive set you have not measured is a list of names you feel bad about, and it produces bad moves: a price change aimed at a firm nobody was choosing between, a marketing budget pointed at a channel your real rivals do not use, a hiring plan benchmarked against a company with a completely different cost structure.
A competitive set is one customer's shortlist, not a directory
You do not compete with everyone in your trade in your county. You compete with the firms a specific customer, holding a specific job, in a specific place, seriously considered alongside you. That is the only definition that predicts anything, because it is the only one attached to a decision that actually happened.
In practice that list is short. Two or three names is normal. A homeowner with water on the floor calls whoever answers; a property manager replacing six units calls the two firms on an approved list; a homeowner planning a replacement in spring gets three quotes because someone told them to get three. Those are three different shortlists with almost no overlap, produced by the same trade in the same town.
The directory list and the shortlist diverge because a customer's search is bounded by time, referral and proximity, not by licensing. The large regional firm with the billboard is in the directory. Whether it is on the shortlist depends entirely on whether it does that job type, at that size, in that postcode, fast enough for that customer.
The unit of analysis is the job type, not the shop
The single most common error here is asking "who are my competitors" as one question. Ask it per job type and the answer splits, because the buying process splits.
- Emergency and no-heat work. Selection is on answer speed and arrival time. Price is checked afterwards, if at all. The competitive set is whoever answered.
- Planned replacement and project work. The customer has time, so they comparison shop, and the set is real and named. This is where a genuine shortlist exists.
- Recurring maintenance and agreements. Selection is on relationship and inertia, and the challenger is usually smaller than you think - a one-van operator with low overhead can hold this work indefinitely.
- Commercial or property-management accounts. Selection runs through an approved-vendor process, insurance certificates and response guarantees. The set is whoever is already on the list, and the real competition is the list itself.
Run the same shop through those four and you will often find four different rivals, only one of whom you had been worrying about.
The three questions that produce the list
You do not need a research budget. You need three questions asked consistently for one quarter, and the discipline to record the answers where they can be counted.
- At the quote, ask who else they are getting a number from. Ask it plainly and without defensiveness: "Are you getting other quotes on this? No problem either way, I just like to know who I am in the room with." Most customers answer honestly. The ones who say "no, you were the referral" have told you something more valuable than a name.
- On a loss, ask who they went with and what decided it. This is the uncomfortable one and it is the highest-yield. Ask a week after the loss, not at the moment of it, and make clear you are not trying to reopen the job.
- On a win, ask why you and not them. The answer is rarely price and almost always something operational - who called back, who explained it, who could come Thursday.
Log four fields per quote: job type, customer type, names mentioned, outcome. Nothing else. A field you will not fill in every time is a field that produces a half-populated table nobody trusts.
Worked example: forty quotes, three segments, three different rivals
A shop logged every quote for one quarter. Forty quotes total: 22 emergency or same-week service calls, 12 planned replacements, 6 maintenance agreements. The three segments behaved nothing alike.
Emergency, 22 quotes. The customer named another firm they had called in 6 of the 22, about 27 percent. In the other 16 they had called one shop, which was this one. Across those 6 mentions, no competitor name appeared more than once. The shop won 18 of the 22, about 82 percent.
Planned replacement, 12 quotes. The customer named at least one other firm in 11 of the 12. Two names accounted for 8 of those 11 mentions; a large regional firm accounted for 2; one other name appeared once. The shop won 5 of the 12, about 42 percent. Split by who was in the room: of the 8 quotes where one of the two frequent names appeared, it won 2 and lost 6. Of the remaining 4 it won 3 and lost 1.
Maintenance agreements, 6 quotes. One name, a two-person operator working out of a single van, appeared in 5 of the 6. The shop won 2, that operator took 3, and 1 customer bought nothing.
Now the finding the owner did not expect. The large regional firm he had spent two years benchmarking against appeared in 2 of 40 quotes overall, 5 percent, and in 2 of 12 replacement quotes, about 17 percent of the only segment it showed up in at all. The two-person van operator he had never once discussed in a management meeting appeared in 5 of 6 maintenance quotes and took 3 of them.
What the map changed
Three decisions moved, and each one had been pointed at the wrong target.
The shop had been planning a price adjustment on replacement work, reasoning from the regional firm's published promotions. Those promotions were shaping the decision in 2 of 12 replacement quotes. The two frequent names were in 8 of 12, and neither of them was advertising a price at all - they were winning on lead time. The correct move was a lead-time move, not a price move.
The maintenance loss was invisible before the count because each instance is small and nobody escalates a lost agreement. Counted, it is 3 of 6 in a quarter, and maintenance is the segment that feeds next year's service calls. That is the loss worth a response.
Emergency work needed nothing. With 16 of 22 customers calling one shop, there is no competitive set to respond to. The constraint there is answer rate and arrival time, and the money that would have gone into a price move went into after-hours coverage instead.
The two errors this fixes
Benchmarking against a firm ten times your size. Its rate structure, its ad spend and its crew count reflect an overhead base you do not carry and a capital structure you do not have. Reading your numbers against it produces a conclusion that you are failing at everything, which is both false and paralysing. If it appears in 5 percent of your quotes, it is a market fact, not your competition.
Ignoring the operator below you. A one-van shop with a home office, no dispatcher and no marketing spend can hold maintenance and small-repair work at a price you cannot match, and it is invisible because it never advertises and never appears in a trade directory. Whether that operator is pricing sustainably or not is a separate question with a separate answer - see related: Competing Against a Shop That Does Not Know Its Own Costs.
The competitor that is not a firm
Two entries belong on the list that no directory contains, and in replacement work they are frequently the largest.
Doing nothing. On a planned replacement the customer's real alternative is usually to defer another season and repair once more. In the count above, 1 of the 6 maintenance customers bought nothing at all, and that outcome is normally filed as a lost quote against whoever else was mentioned. It should be filed separately, because it responds to completely different things: financing, a staged scope, a clear statement of what failure looks like if they wait. A price cut aimed at a rival does not win a deferral, and shops routinely discount into a decision that was never a comparison.
The generalist. A handyman, a maintenance man on a property manager's own payroll, or the customer's brother-in-law will take small repair work and some of your recurring work at a rate no licensed shop can meet, and they do not appear in your trade's directory because they are not in your trade. Count them when they are named. Where they are taking whole job types, that is a scope decision for you, not a pricing one.
Record both as distinct outcomes rather than folding them into "lost". That quarter produced fifteen non-wins: fourteen jobs that went to somebody else, and one customer who bought nothing from anybody. Folded together they read as fifteen competitive losses, and the reflex is a price move. Split, the fourteen send you to lead time and a one-van operator, and the one sends you to financing and scope. The folded number tells you nothing about which of those you actually have.
Where you have no real competition, and what that is worth
The segments with no shortlist are the most valuable output of the exercise and the easiest to miss, because nothing is going wrong in them.
Where the customer called one firm, you are not being priced against anything. That is where margin is defensible, where a scope increase is accepted without a second quote, and where an annual increase goes through on a sentence rather than a negotiation. It is also fragile in a specific way: it usually rests on response, referral or a single relationship, and it disappears the first time you cannot get there.
So treat those segments as an asset with a maintenance requirement. Know which ones they are, know what holds them - the after-hours rota, the two referring plumbers, the property manager who has your cell number - and protect that specific mechanism rather than the segment in general.
How to know you got this right
You have a real competitive map when three things are true. Every name on it came out of a customer's mouth rather than a search result. Each job type has its own list, and at least two of those lists differ. And you can say, per segment, what the named rivals are actually winning on - lead time, price, an approved-vendor listing, a relationship - rather than assuming price by default.
The failure mode is a map built from what you see rather than what customers say: the firms whose vans you notice, whose ads you see, whose owner you know. That map is biased toward whoever markets hardest, which is exactly uncorrelated with who is taking your work. If your list is the same list you could have written before you started counting, you have recorded your assumptions, not your market. Re-run the count annually, and immediately after any change in your own service area, because the shortlist is bounded by drive time and moves when you do.
References
- See related: Competing Against a Shop That Does Not Know Its Own Costs, Competitive Positioning, The Comparison Trap: Your Shop vs Theirs
- See related: Why Pricing Against Your Competition Is a Trap, Benchmarking Against Your Own History Not Someone Else's Business
- U.S. Small Business Administration, market research and competitive analysis guidance for small business