GrowthStack Advisory / Hands-on experience / Outbound sequence structure
How to structure a B2B outbound sequence
Seven touches is a reasonable starting shape for a B2B outbound sequence: a hook, a proof point, an angle change, then variations. But the count matters far less than whether each touch says something new. A three touch sequence with three distinct angles beats a twelve touch sequence that repeats the same pitch louder. Email, LinkedIn and phone should be designed as one motion rather than three disconnected efforts, and the sequence is the last thing to build, not the first: deliverability, the account list and the messaging framework all come before a single send. What separates sequences that work from sequences that generate activity is not cleverness, it is discipline: a fixed touch pattern, a written exit rule, one variable changed at a time, and reply analysis across the whole campaign rather than reactions to individual replies. Shobhit Gupta has built this motion in house at Locus, GoComet and Landmark Group, and as a consultant for Easyship, ConvertCart and nuvista.ai.
Written and maintained by Shobhit Gupta, Founder and Principal GTM Consultant at GrowthStack Advisory. Ten years running sales development and revenue operations at Locus, GoComet and Landmark Group's Logistiq division. Published
The touch count is the wrong question
Almost every question about sequence design arrives as a number. How many emails. How many days. How many touches before giving up. The number is the easiest thing to copy from someone else's playbook and the least useful thing to get right.
Around seven touches is a reasonable starting shape. It is enough room to make an argument, support it, and try a second angle before the account has decided. But a three touch sequence with three distinct angles beats a twelve touch sequence that repeats the same pitch louder, because the second one is not really twelve touches. It is one touch sent twelve times, and the buyer read it the first time.
The test that actually matters: for each touch, can you say what it adds that the previous one did not? If not, the sequence is padded rather than designed, and adding touches makes it worse rather than longer.
What each touch is for
A sequence that works has a shape, and each position in it does a different job.
- The hook. The reason this account specifically, said in the buyer's language rather than yours. If the opening line would work unchanged on a different company, it is not a hook.
- The proof point. One claim, with evidence mapped to it. Messaging is written per segment and per persona rather than per campaign, so the proof is the one that segment finds credible.
- The angle change. A different problem, not the same pitch louder. This is the touch most sequences skip, and skipping it is why touch five reads like touch two.
- The variations. Short, direct, and willing to close the loop. A sequence that never offers the account a way out reads as pressure rather than persistence.
Every claim in that sequence needs a proof point mapped to it. This is where the messaging framework does its work: narratives written per persona, with evidence attached, rather than copy written per campaign. More on that order in ICP, persona and messaging.
Three channels, one motion
Email, LinkedIn and phone should be designed as a single sequence that happens to use three channels, not three sequences that happen to share a list. Run separately, they produce three weak impressions instead of one coherent one, and they make attribution impossible, because nobody can say which channel moved the account.
- Email carries the argument. It is the only channel with room for a problem statement and a proof point in the same message.
- LinkedIn carries the recognition. It makes the name familiar before the call, which changes the call from cold to warm without changing anything about the call itself.
- Phone carries the conversation. Everything the first two channels cannot do, which is most of what actually qualifies an account.
The reason this matters commercially: a coordinated motion is the part an SDR agency is least likely to give you, because it requires the three channels to be owned by the same person with the same context. That distinction is set out in SDR agency vs GTM consultant vs in-house.
The five disciplines that decide the outcome
Sequence copy gets the attention. These get the results.
- Deliverability before volume. Separate sending domains, SPF, DKIM and DMARC configured, mailboxes warmed, rate limits set. A perfect email in spam converts at zero. The full checklist is in cold email deliverability.
- A fixed touch pattern, written down. The number of touches, the channels and the interval, set in advance and held to, so follow up does not depend on who remembers.
- A defined exit rule. A written rule for when to stop, so reps are not guessing. A lead worked once and abandoned is not a bad lead, it is an unworked lead.
- One variable at a time. From launch onwards. Change the list, the subject line and the cadence in the same week and you will learn nothing when the number moves.
- Reply analysis across the campaign. Read what came back and name the pattern, rather than reacting to the loudest individual reply. This is the step that tells you whether the problem is the targeting or the message.
Speed is a different problem from cadence
Outbound cadence and inbound response speed get confused, and they are not the same lever. Cadence is how you work an account that has not asked to hear from you. Speed applies to the account that just raised its hand, and there the evidence is stronger than for anything else in sales development.
Contacting a web lead within five minutes rather than thirty made a large difference to qualification odds (MarketingSherpa, 2007), and an audit of 2,241 US companies found the average first response took 42 hours, with 23% never replying at all (Harvard Business Review, 2011). A follow up rule that survives a busy week beats a cleverer email. The full argument is in improving MQL to SQL conversion.
Personalisation, and what it is not
Inserting a company name into a template is not personalisation, and buyers read it as exactly what it is. Personalisation means the first line is specific because the research behind it was real.
The practical constraint is time. A first line that costs twenty minutes of research per account does not survive contact with a quota, which is why most personalisation programmes quietly become merge fields within a month. This is the part of the workflow where AI earns its place: account research, personalisation at volume, reply analysis and lead prioritisation, with human review before anything sends. That division is set out in AI as a layer, not another tool.
Build the sequence last
The sequence is the visible part, which is why teams start there. It should be close to the last thing built. The ICP and the named account list come first, then the messaging framework, then the sending infrastructure, and only then the sequence that carries all three.
Sending before the list exists wastes the diagnostic value of the first month, and sending before authentication is configured means the first month's data is measuring deliverability rather than message quality. The week by week order is in the first 90 days of outbound.
Where this has been run
This is the motion behind the named client work rather than a general framework. At Easyship, tightening inbound handling and qualification moved the inbound qualification rate from 15% to 25% inside the first month. At ConvertCart, meeting no-shows fell from 44% to 15% once follow up discipline and the confirmation flow were fixed, with LinkedIn added as a second channel. For nuvista.ai, the whole motion was built from zero.
In house, the same structure ran at Locus across the United States, Europe, the Middle East and Asia, where MQL to SQL conversion moved from 10% to 25%, and at GoComet, where output went from 2 to 6 SQLs per rep per month as the team scaled from 10 to 23. The full case studies set out what changed in each.
References
Frequently asked questions
How many touches should a B2B outbound sequence have?+
Around seven is a reasonable starting shape: a hook, a proof point, an angle change, then variations. But the count matters far less than whether each touch says something new. A three touch sequence with three distinct angles beats a twelve touch sequence that repeats the same pitch louder. If you cannot say what each touch adds that the previous one did not, the sequence is padded rather than designed.
Should email, LinkedIn and calling be separate sequences?+
No. They should be one motion. Three disconnected efforts against the same account produce three weak impressions rather than one coherent one, and they make attribution impossible because nobody can say which channel moved the account. Design the sequence as a single pattern that happens to use three channels.
What should the interval between touches be?+
That is set per engagement rather than by a universal template, because it depends on the buying cycle, the segment and how the account entered the sequence. What matters more is that the interval is written down and held to, so a rep is not choosing it in the moment. An unworked lead and a bad lead look identical in the CRM.
How fast should inbound leads be contacted?+
In minutes, not days, and this is the strongest evidence in sales development. Contacting a web lead within five minutes rather than thirty made a large difference to qualification odds (MarketingSherpa, 2007), and an audit of 2,241 US companies found the average first response took 42 hours with 23% never replying at all (Harvard Business Review, 2011). Outbound sequences and inbound response are different problems and the speed rule applies to the second.
When should a rep stop working an account?+
When the written exit rule says so. Without one, reps guess, and the guess is usually either abandoning after two touches or working a dead account for months. Both are expensive and only one of them is visible in the CRM.
Why does deliverability come before the sequence?+
Because a perfect message that lands in spam converts at zero, and it is the cheapest failure to prevent. Authentication, sending domains, mailbox warmup and list hygiene go in before volume. A deliverability problem reads exactly like a copy problem, which is how teams end up rewriting subject lines for a month while the real fault sits in DNS.
How do you test a sequence without losing the ability to attribute results?+
Change one variable at a time. If the list, the subject line and the cadence all change in the same week, a number that moves has no known cause and the sequence cannot be improved, only replaced. Copy, sequence, sending infrastructure and list quality are separated and tested in order.
Does personalisation actually matter at volume?+
It matters when the input is real research rather than a merge field. The goal is a first line that is specific without costing twenty minutes per account, which is where AI genuinely helps: account research, personalisation, reply analysis and prioritisation, with human review before send. Inserting a company name into a template is not personalisation and buyers read it as exactly what it is.
Related reading
Engagements start at $2,000 per month, after an optional $750 GTM diagnostic. See how pricing works, what is not included, and the notice terms.
Tell us where the funnel leaks
Thirty minutes is enough to tell you whether the problem is the sequence, the list, or the handling after the reply.