How to Decide Which Service to Repeat and Which to Drop

Write your services in a column. If that column is longer than the number of jobs you completed last month, you are looking at a list of things you are willing to attempt rather than a record of what you do, and it is making every enquiry harder to answer and every advert harder to write.

The way out is not to pick a favorite. It is to score the list against your own records and let the scoring make the decision while you are calm.

The four axes to score each service on

Give every service on the list a low, medium or high on each of these. No math required.

Effort. What the job costs you door to door — travel, setup, the work itself, cleanup, and the chasing afterwards. People underrate this because they count the hours on site and forget the hour getting there and the three weeks getting paid. If your sense of a job's real cost is fuzzy, working out what a side hustle actually costs you in time is the measurement to do before scoring anything.

Repeatability. Could you do this again next week, the same way, without thinking hard about it? A service that requires improvization every time cannot get faster.

Demand. How many enquiries arrived for it without you pushing. Count from records, not memory — memory over-weights the recent and the dramatic.

Whether it leads anywhere. Does finishing it open a door — to a larger job, a return visit, an introduction — or does it simply end?

Score honestly, then look at the shape rather than any single column. High effort with low demand is the easiest cut you will ever make.

The service you enjoy and the service that carries you

The service you love is on your list because of how it feels to do it. That is a real thing and worth protecting, but it is not evidence that anyone wants it.

The one that pays is sometimes the one you find dull, and there is a mechanical reason for that: dull to you means automatic to you, and automatic means fast. Boredom is what competence feels like from the inside.

So branch. If the enjoyable one also scores well on demand, keep it and build on it — that is the easy case. If it scores badly, keep doing it, but stop advertising it and stop letting it occupy space in how you describe yourself. It becomes the thing you take on when someone asks, not the thing you sell.

And if the dull one has demand and low effort, that is your candidate. Make it the thing you get quicker at.

Loss leaders that earn their place, and ones that only feel like they do

A loss leader earns its place when your own records show a path from it to the profitable work, with names you can point at.

A fake one survives for softer reasons. It is how you started. A competitor offers it. It makes the list look complete. Or one person, once, bought the bigger job afterwards, and that single event has been carrying the whole justification ever since.

The test is short. Take every customer who bought the cheap thing and count how many went on to buy the next thing. If that count is zero, it was never a loss leader, it was a discount. If it is one, you have an anecdote. If several people made the jump, you have found something worth keeping and worth pointing deliberately at the next step rather than hoping.

The other honest use for a cheap service is as a way for people to try you at low risk. That is legitimate, but it is only working if the people who try you come back. Same test, same records.

Dropping a service without dropping the customer who bought it

Tell people before they ask, not when they ask. A customer who books and then gets told no remembers the refusal; a customer told in advance remembers the notice.

Give one plain reason and stop there. I have narrowed down to the guttering so I can do it properly is enough. Do not apologize at length and do not explain your business strategy, because a long explanation invites negotiation.

Hand them somewhere to go. A name you would actually use, not a search suggestion. If work is already booked in the category, finish it — dropping a service does not mean breaking commitments.

Then keep them on your list for the services you kept, and say so explicitly, because otherwise a dropped service reads as you disappearing. The hard part comes when a good customer asks for the dropped thing as a favor, which will happen, and saying no to work that breaks your repeatable system is where that conversation gets handled properly.

The re-test: what would bring it back

Write the conditions for return now, while you still remember why you dropped it. Without them you will re-add it during the first quiet week out of pure anxiety, and the list grows back.

Good conditions are observable, not felt. Three unprompted enquiries for it. A method that removes whatever made it slow. One customer willing to pay the price that would make the effort score acceptable. A tool you do not currently own.

Put a review date on it, chosen by you, and put the note somewhere you will actually find it — the same place the scoring lives. Then leave it alone until then.

When the list is finally short, two things follow. The survivors want packaging into something people can buy without a conversation, which is building a repeatable offer instead of custom work. And a short list is what makes capacity questions solvable at all, so growing without hiring anyone becomes worth reading once you are doing one thing repeatedly.

Wait until the record is long enough

If you have completed a handful of jobs, you do not have data. You have moods. Scoring four axes across three afternoons produces a confident answer that is wrong, and the confidence is the dangerous part, because you will act on it and cut the thing that was about to work.

Stay broad for now. Keep saying yes to a range. Write down every enquiry that comes in, including the ones you turned down and the ones that went nowhere, because that record is the raw material this whole exercise runs on and you cannot reconstruct it later.

Come back when the same request has arrived enough times that you recognize it before the person finishes the sentence. And do not score your list during a bad week — a run of difficult jobs makes everything look unprofitable. Score it against the record, not against how last Tuesday felt.