How do I build a repeatable sales system instead of a lucky quarter?

What is the difference between a spike and a system in sales?

A spike is a sales result you cannot repeat because you cannot name what produced it. A system is a result you can repeat because you can. That is the whole definition. Everything else on this page is how you move a result from the first column to the second.

The uncomfortable part is that the two can post the same number. Same attainment, same bar on the chart, same congratulations in the same meeting. The dashboard cannot tell them apart, because it records what happened and never why. The difference lives in one place: whether anyone in the building can say, in a plain sentence, what produced the result. A quarter nobody can explain is a liability, however big it is. You will plan against it anyway, and that is where the damage starts.

If you want to test whether a specific great quarter was a spike, the diagnosis is here. This page covers the other half, what to build once you know.

The build is Isolate, Refine, Compound (I.R.C.). Three moves that turn a sales result you cannot explain into a behavior you can run again. They run in order, because each one needs the one before it.

How do you isolate what actually won a deal?

Use the one-sentence test. Take your last five wins. For each one, write a single sentence that names a mechanical move another rep could repeat on Monday. A move, with a verb in it. Not a trait, not a feeling.

Words like chemistry, instinct, relationship, timing and territory are closed boxes. They sound like explanations and contain no behavior. The site calls them the Lazy Vocabulary, and the test is simple: if the rep at the next desk could not do what your sentence says, the sentence failed.

Here is how it fails, and how it gets fixed, on one of my own deals. It was the renewal of my company’s largest customer. I had twelve years on the account. Leadership on the customer side had cooled, and a competitor was holding a real plan to replace us. It closed in three days, against best-and-final positions both sides had called fixed.

My first explanation went like this: I had a gut feeling I could make a deal, and the relationship carried it. I knew both sides well enough to tell what was real from what was posturing. True, probably. Also useless. There is no behavior in it, and nothing the rep next to me could learn.

So I went back through it and wrote what I actually did. Five moves. I asked for more time, three days, having asked for five. I went around the formal channel and called my main contact directly. I went back to my own leadership and separated what we truly needed from positions we had taken in response to the customer’s positions. I gave them the weekend to feel the gravity, a deliberate pause. And I asked the customer’s VP for a call and asked one question: what was his long-term goal with the contract?

That was better, and still a list. The mechanism came on the third pass. I had treated my own leadership as a discovery target, running the same needs-analysis sequence I train reps to run on prospects, separating stated position from real need. I had treated time as a negotiated variable instead of a fixed one. And the VP question was the opening move of discovery, applied at the executive level. That is the Third Pass, and it produced the sentence I could not write on the first pass.

Notice that nothing in those three passes was new information. I knew all of it the day the deal closed. What was new was the sentence. If you cannot name it, you cannot repeat it. If you cannot repeat it, you do not own it.

One caution before you run the test on your team. If the win belongs to your top rep, do not ask them how they did it. They are rarely doing what they say they are doing. Watch the calls and write the sentence yourself, which is the argument in why your top rep can’t be replicated.

How long does it take to refine a sales behavior?

Pick one behavior from the list. Just one. Run it for one quarter, and start on deals where a miss is cheap before you use it on the deal you cannot afford to lose. Most refinements take about a quarter to settle.

Expect the Friction Weeks. These are the two to twelve weeks where the new behavior is installed and not yet producing anything. The rep can look worse in this stretch, because a rep who is thinking about a move is slower at it than a rep who is running on habit. Reps abandon refinements that were going to work after three weeks of bad luck on outcomes. It is the most common way a good change dies.

So judge the behavior. The weekly question is not "did it close." It is "did I do the thing I said I would do this week." Three weeks of results are mostly weather. Three weeks of behavior are a record you control.

What actually compounds?

Outcomes do not compound smoothly. They lurch. Quarters spike and crater for reasons that sit outside the work: a customer merges, a competitor stumbles, a budget shows up early. What compounds is the stack of named behaviors, each one isolated, refined and left running while the next is added.

That line is I.R.C. in three sentences. Find is Isolate. Improve is Refine. Repeat is Compound. And the last clause is the argument: the results become unavoidable because they sit downstream of something you run every week.

Here is what the difference looks like across a team. Take two lines over eight quarters that reach the same total. One is jagged. One is smooth. The jagged line is carried by a handful of deals, a handful of reps and a handful of relationships, and when you remove any one of them, the line flattens. The smooth line is carried by behaviors every rep runs, so there is nothing to remove.

Same total, different risk. The danger sits in what gets built on top of the jagged line. Hiring plans, territory cuts, comp design and forecast calls all get built on that pattern, because it looks like the pattern of a healthy team. When it does not repeat, the decisions did not go wrong. The pattern they were built on did.

A named behavior pays out on a schedule you can predict. Sourcing is the clean example. What your team starts sourcing now shows up as closed business a quarter later, which is the whole argument of Q4 is decided in August. You can see that only because sourcing is a behavior with a name.

Run it this week.

1. Pull five wins. Take your last five closed-won deals, from you or from your team.
2. Write one sentence each. Name one mechanical move another rep could repeat. Cross out any sentence built on chemistry, instinct, relationship, timing or territory, and rewrite it with a verb.
3. Find the repeats. Look across the five sentences for the two or three behaviors that show up more than once. Those are your candidates.
4. Choose one. Pick the one behavior you will refine this quarter, and pick the low-stakes deals where you will run it first.
5. Set the weekly question. Write it down: did I do the thing I said I would do this week. Ask it every Friday, and do not change the behavior for at least three weeks, however the outcomes look.

You do not need a better quarter. You need a quarter you can explain, and then another one. Money will still arrive unevenly, and some of it will be luck, which is fine. Take it, and do not confuse it with craft.

See what you do. Change one thing. Let the years multiply it.