A prospecting cadence is the planned run of touches (emails, calls, voicemails, LinkedIn messages) an SDR or AE makes to one contact over two to three weeks before stopping. Building one takes about an hour for the template and ten minutes per account to fill in, and the rule that decides whether it works is simple: every touch has to give the buyer a new reason to reply.
The cadence carries the email and the call covered elsewhere in our SDR coaching guide. This page builds the sequence around them.
What a prospecting cadence has to do
A prospecting cadence has to make one argument to one buyer, in several pieces, across channels. Each touch adds a piece. When a touch adds nothing, the buyer reads it as pressure and the sequence stops working from that point on.
- Start from the problem
- Decide who is in it
- Set the length and the stop rule
- Give every touch a job
- Write touch one under the cold email rules
- Carry the problem across channels
- Read the whole sequence before launch
- Review weekly, one touch at a time
Most cadences fail the same way: the same email five times, reworded, with rising urgency ("just bumping this", "circling back", "one last try"). A rep who follows that cadence is following instructions. The fix is in the template. Our outbound use case puts it as a sequence "where every touch adds something", so that touch four still gives the buyer a reason to answer.
Eight steps to a cadence that builds
1. Start from the problem
Write the one problem the cadence is about, from the account research: the person, the problem, the evidence. Good looks like: one sentence that every touch can refer back to. A cadence with no stated problem becomes a product announcement repeated five times.
2. Decide who is in it
Name every person the cadence reaches, and decide whether they get the same touches. Teams on Vortex Prospecting reach the economic buyer, the user buyer and the technical buyer in the first wave, so the deal never rests on one contact. Good looks like: at least two people per account, each with a reason to care.
3. Set the length and the stop rule
Our operating default is five touches over about three weeks for a first cadence, with a clear last touch. Longer cadences work when each extra touch has a job. Good looks like: a written stop rule (after the last touch, the contact rests for a quarter unless a new trigger appears).
4. Give every touch a job
For each touch, write the reason to reply it adds. Use the table below. Test: if you cannot name what touch three adds that touch two did not, cut it or change it.
| Reason a touch can add | Example |
|---|---|
| The problem, stated first | "Three night dispatch roles posted this month" |
| A new angle on the same problem | The new depot doubles the handovers |
| Proof from a peer | What changed at a similar company, with one specific |
| A fresh trigger | A new hire, a launch, a quarter-end announcement |
| A different channel | The call that refers to the email |
| A different person | Copying the operations director with a reason that fits their role |
5. Write touch one under the cold email rules
Touch one sets the argument. Our house rules: under 100 words, the buyer's problem before the product, an ask they can answer in one word, and no link. The cold email how-to covers it step by step.
6. Carry the problem across channels
A call or voicemail touch should name the problem the email raised. "I wrote to you about the night handover" makes the call the second half of the email. A call that starts a new topic resets the argument. The cold call script template has a fifteen-second voicemail built this way.
7. Read the whole sequence before launch
Read all five touches in order, as the buyer would. Mark each one "builds" (adds a reason) or "nags" (repeats one). Good looks like: five builds. A manager should do this read for every new template and for a sample of each rep's filled-in sequences.
8. Review weekly, one touch at a time
Each week, look at where replies come from and where they stop. Find the first touch after which nothing comes back, read it against the touch before it, and rewrite that touch. Good looks like: one change per week, tested on the next batch. Changing everything at once tells you nothing.
A worked five-touch cadence
Illustrative. The company, people and details are fictional.
Maya sells dispatch software. The problem from her research: the night handover at Marlby Freight is losing information as the company adds depots. The cadence reaches Sam (head of dispatch) and, from touch four, Lee (operations director).
| Touch | Day | Channel | Job | The line that carries it |
|---|---|---|---|---|
| 1 | 1 | Email to Sam | The problem, stated first | "Marlby has posted three night dispatch roles this month... Want the three things we saw? Yes or no is fine." |
| 2 | 3 | Call and voicemail to Sam | Different channel, same problem | "Sam, Maya at Depotline. I wrote about the night handover. Is that a staffing question at your end, or a handover one?" |
| 3 | 7 | Email to Sam | A new angle | "The new depot opening doubles the shift handovers your team runs each night." |
| 4 | 11 | LinkedIn note to Lee | A different person | "Lee, your post about the new depot made me think of the handover between night shifts. Is that on your list or Sam's?" |
| 5 | 16 | Email to Sam | Proof from a peer, and the last touch | One specific from a similar company, then: "If it's not a priority, I'll leave it there. Want the note?" |
Read in order, each touch builds. Touch two carries the email's problem into the call. Touch three moves the argument forward with the depot. Touch four brings in a second person with a reason that fits their role. Touch five adds proof and ends the sequence with a clear last ask.
The version Maya first wrote had touch three as "Just following up on my last email" and touch five as a calendar link. Both nagged.
Where cadences go wrong
The same email five times. Rewording does not add a reason. Each touch needs a job from the table.
Channels without an argument. A cadence that alternates email, call and LinkedIn because the template says so, with each touch on a different topic, gives the buyer five unconnected interruptions.
The calendar link early. Asking a buyer to book time before they have replied once asks for the meeting before the reply. Keep the ask small until they answer.
One contact. A cadence that reaches one person stops the day that person is on leave or changes roles. Reach two or three people with reasons that fit each of them.
No review. A cadence launched and never read again keeps its weakest touch forever. One touch, reviewed weekly, is enough to improve it.
Reading every sequence before it runs
Step 7 is the one most teams skip, because a manager with eight SDRs cannot read every filled-in sequence before it launches. That is where a nagging touch survives.
Our AI sales experts work in the team's prospecting method (for SDRs): Keenan for teams on Gap Prospecting, Val for teams on Vortex Prospecting, and Caty for a house method. The expert scores each email touch against the method and names the one that stops building; the rep talks the rewrite through on a call, or types when they cannot talk. Bryce, the expert who coaches managers, gives the SDR manager the read across the team's calls and deals.
Start by rewriting touch one, and talk to us about the sequence you were about to launch. For the definition of the term, see what is a prospecting cadence.
FAQ
How many touches should a prospecting cadence have?
Five touches over about three weeks is a sound default for a first outbound cadence; that is our operating guidance. More touches work if each one adds a new reason to reply. The number matters less than the test: if a touch only repeats the last one, it should be cut or rewritten, whatever the total.
How long should a sales cadence last?
Most outbound cadences run two to three weeks. Space the early touches closer together and the later ones further apart, and set a stop rule: after the last touch, the contact rests until a new trigger, such as a hire or launch, gives a fresh reason to write. Endless follow-up turns a cadence into noise.
What channels should a prospecting cadence use?
Use email, phone and LinkedIn, with each channel carrying the same problem. A call should refer to the email before it, and a LinkedIn note can bring in a second person. The channel mix matters less than whether each touch adds a reason to reply that the last one did not.
What should a follow-up email say?
A follow-up email should add something the first email did not: a new angle on the problem, a specific from a peer, or a fresh trigger at the account. "Just following up" or "bumping this to the top of your inbox" adds nothing, and the buyer reads it as pressure. Keep it short and keep the ask small.
How do you know which touch in a cadence is failing?
Look at where replies come from and find the first touch after which nothing comes back. Read that touch against the one before it: usually it repeats the earlier message. Rewrite that touch only, run it on the next batch, and check again the following week.
What is the difference between a cadence and a sequence?
A cadence and a sequence mean the same thing in most sales teams: a planned series of touches to one contact across channels and days. Some tools say "sequence" and some say "cadence". Separately, "revenue cadence" or "operating cadence" means a team's rhythm of forecast and pipeline meetings, which is a different topic.