You've seen this play out: the cadence ran for six weeks, meetings came in below target, and in the Monday pipeline review someone says "let's rewrite the emails" or "let's add more calls." The change goes live Tuesday, and afterward nobody can tell whether results improved because of it or in spite of it.
Changing the cadence isn't the problem. Changing it without knowing what was broken is. And the most honest way to find out isn't the rolled-up reply rate report. It's the touch history for each account: the actual sequence of what happened with every company you reached out to.
Why the aggregate number misleads you
A reply rate of, say, 4% can hide completely different situations. Maybe half the accounts never got the third touch. Maybe every reply came from a single segment. Maybe phone is converting well but was only used on 20% of accounts because valid numbers were missing.
The aggregate tells you how much. The account-level history tells you how. And to decide what to change, you need the how.
What to look for in the touch history
In your dashboard, pull a sample of accounts and read each one's timeline. You don't need hundreds; a sample of 30 to 50 accounts, mixing ones that converted, ones that replied with a no, and ones that went silent, usually surfaces the patterns. For each one, check:
- Executed vs. planned cadence: did every planned touch actually happen, at the right intervals? Accounts that stopped at touch 2 didn't test the cadence; they tested half of it.
- Contacts per account: did you reach just one person or the buying committee? In B2B tech sales, single-threading with one decision-maker tends to cap results hard.
- Channels actually used: was the LinkedIn request accepted? Did the call reach a person or a main line? Did the email bounce?
- Which touch generated the reply: if positive replies cluster around touches 4 to 6, shortening the cadence to save effort would be a mistake.
- What the reply said: "bad timing," "I'm not the right person," and "we already have a vendor" point to different problems.
Sorting out the four types of problems
Once you've read through the sample, the diagnosis almost always falls into one of these buckets. Each one calls for a different fix.
1. List problem
Signs: lots of "not the right person," wrong titles, companies outside your profile, bounced emails. Rewriting copy won't fix any of this. The fix lives in your ICP and contact research.
2. Execution problem
Signs: late touches, skipped steps, calls never made, accounts dropped midway. The cadence may be fine and simply never ran. Before changing it, make sure it gets executed as designed for one full cycle.
3. Messaging problem
Signs: right list, full execution, but near-total silence or replies like "I don't get what you do." Now it's worth reworking the copy, the pain angle, or the first-touch offer.
4. Channel or timing problem
Signs: one channel drives all the replies while another produces nothing, or replies only show up on late touches. The fix is rebalancing channels and adjusting intervals, not necessarily changing the content.
Pro tip: before making any change, write down in one sentence which of the four problems you found and what evidence in the touch history backs it up. If you can't point to at least five accounts that show the pattern, you're still giving an opinion, not a diagnosis.
How to change without losing your baseline
With the diagnosis in hand, the rule is to change one variable at a time. If both the list and the messaging were weak, fix the list first: it contaminates everything else. Changing everything at once wipes out any chance to learn.
- Document the current version: save the cadence as it was, with channels, intervals, and copy.
- Define the change and the hypothesis: "adding a call at touch 3 should increase replies from accounts that haven't responded to email."
- Run it on a comparable batch: same account profile, same time period, so you don't credit the cadence for an effect that came from the segment.
- Let the cycle finish: a 15-day cadence needs at least that long, plus a window for late replies, before you judge it.
- Go back to the account-level history: confirm the change was actually executed and see where replies showed up.
What the audit usually reveals
At Nord, when we run this exercise using the dashboard history, we often find the cadence wasn't the villain. Frequently the lift came from multithreading into more people per account or pulling companies out of the batch that never should have been there. That kind of fine-tuning, more than reinventing the messaging, is what sustained results like 32 meetings a month for a logistics company in Spain.
The opposite is common too: the team thinks email isn't working, but the history shows the positive replies came from the fourth email, right after two unanswered calls. Cutting that email would have tanked the results.
Next steps
The next time someone suggests changing the cadence, ask for 30 minutes before the decision. Open the touch history for a sample of accounts, classify the problem as list, execution, messaging, or channel, and only then pick a single variable to change. Document the hypothesis and review it when the cycle closes. It's a simple habit that turns cadence tweaks from guesswork into compounding learning.





