A Cold Email Sequence Built Around the Age of the Signal
A cold email sequence usually runs on a calendar that starts the day you enrol someone. When the reason for writing is a dated store event, the sequence has a second clock — how old the signal is — and that one decides more.
Agencies and software companies selling to ecommerce stores
- →A cold email sequence runs on two clocks: days since you enrolled the prospect, and how old the triggering event is. Most sequence templates only track the first.
- →Our own buying signal catalogue states the position without hedging — we surface events that are days old rather than weeks, and a three-week-old signal is "noise dressed up as an opportunity". Retrieved 22 September 2026.
- →Stores are re-scraped weekly, so a signal's age is accurate to about seven days, not to the hour. Sequences should be designed around that resolution rather than around real-time detection we do not offer.
- →Signal pools differ by more than an order of magnitude — roughly 610,000 stores with an email tool installed against 31,000 whose active ad count is rising, as of September 2026. Enrolment rate, and therefore the decay window, follows from which signal you pick.
- →We are not going to give you the optimal step count, the right decay window, or a reply rate for age-aware sequencing against calendar sequencing. We have not measured them, and a figure from our campaigns is not a forecast for yours.
What a cold email sequence is actually scheduling
A cold email sequence schedules two things at once, and most templates only admit to one. The first is your clock: day 1, day 3, day 7, counted from the moment you enrolled the prospect. The second is the age of whatever made you enrol them. When the trigger is a dated event, those two clocks disagree almost immediately.
The disclosure first. We build Keaz Signals, which sells ecommerce store data with dated buying signals attached. We sell the second clock. We are a competitor to most of the tools that sell the first one, and we have an obvious interest in you concluding that enrolment date is a weak scheduling input. Read the rest with that in mind.
What follows is a reading of our own product behaviour and our own published position, retrieved 22 September 2026. There is no benchmark study here and no reply-rate table, for reasons set out near the end.
What the day-1, day-3, day-7 template assumes
The standard sequence assumes every prospect enters in the same state. Enrolment is treated as the start of the story: step one goes out immediately, and the later steps are spaced off it. That works when the reason for writing is timeless — you sell something, they might want it, today is as good as any other day.
It stops working the moment the reason for writing has a date on it. Two stores can enter the same sequence on the same Monday: one because it installed Klaviyo on Friday, one because it installed Klaviyo five weeks ago and you only just got round to building the list. The sequence treats them identically. The first email is the same email. But the Friday store is mid-setup with the flows unbuilt, and the five-week store has either finished or given up, and no amount of step spacing recovers that difference.
This is a separate question from whether any given follow-up has something new to say, which we have written about in cold email follow-up, timed by the signal not the calendar. The problem here sits upstream of the follow-up. It is about who was allowed into the sequence at all, and in what state they arrived.
Signal age is a different axis from sequence step
Sequence step counts your messages. Signal age counts the store's distance from the event. The two are independent, and collapsing them into one number is where age-blind sequences go wrong.
Our position on this is published on the buying signal catalogue, and it is not hedged. Every signal is stored with the date it was last seen; we surface events that are days old, not weeks; and a three-week-old signal is described there as "noise dressed up as an opportunity, and treating it as one is how outreach starts to feel like spam again". Retrieved 22 September 2026.
That is a statement about the data, not about your copy. It means the useful question at enrolment is not "is this store a fit" but "is this store a fit, and is the event still current". A sequence that cannot ask the second question will keep sending confident, well-written first emails about things that finished happening last month — and it will read as a form letter, because functionally that is what it is.
What a weekly re-scrape does and does not give you
Stores in our pool are re-scraped weekly. That sets the resolution of every age figure we can give you: a signal's age is accurate to about a week, not to the hour. We are not a webhook on somebody's Shopify admin, and a sequence designed as though we were will be built around precision we do not have.
As of September 2026 we watch 4.1 million Shopify and WooCommerce stores across Europe and North America. The per-signal pools published on the signal catalogue on 22 September 2026 differ by more than an order of magnitude: roughly 610,000 stores with an email marketing tool installed, 370,000 with new Meta ads running, 128,000 with a product launch, 96,000 with store and site changes, 80,600 with newsletter activity, 34,800 with social growth, and 31,000 whose active ad count is rising. Those figures move, which is why they are dated.
The consequence for sequence design is enrolment rate. A sequence keyed to a rare event fills slowly and can afford a long decay window. One keyed to a common event fills faster than most teams can work it, and the window has to be short — otherwise the queue ages while it waits, and you are back to mailing five-week-old news with a fresh timestamp on it.
Entry, decay and exit: the three decisions age changes
An age-aware sequence makes three decisions a calendar sequence never has to. Entry: the signal must be current when you enrol, not when you built the list. Decay: the sequence expires, and the remaining steps are cancelled rather than sent. Exit: a newer signal on the same store supersedes the running sequence instead of stacking on it.
Decay is the one that gets skipped. Cancelling queued messages feels like waste, and the cost of sending them is invisible — it lands on the recipient and on your sending domain, not on your dashboard.
Exit matters because stores generate more than one signal. A store that installed Klaviyo in March and launched a product last week is two conversations, and only the second is live. Running both at once sends two unrelated emails from one sender in a week. One signal per sequence is the rule that falls out of this.
Our AI sales agent writes each message per store from the signal plus your knowledge base. Patterns per event type are in cold email templates by signal type.
The numbers we are not going to give you
We are not going to tell you the optimal decay window. It is the number this article most obviously points toward — seven days, fourteen, thirty — and we have not measured it. The honest answer is that it almost certainly differs by signal: an unfinished email-tool setup and a product launch do not go stale at the same rate, and we have not measured either one.
We are also not going to put a reply-rate number on age-aware sequencing. Our own site publishes a figure from our own campaigns, dated August 2026; it came from our offers, sent to our segments, on our domains. It is not a forecast for yours, and presenting it as one would be the most persuasive dishonest thing in this article. A reply rate is a property of your offer, your segment and your sending domains together, and a vendor controls at most one of those. The optimal step count is the same kind of question, and we do not have that number either.
And nothing here is legal advice about who you may contact or what you may send them. We are GDPR-conscious by design and we will not claim more than that; responsibility for what you send stays with you.
Where a calendar sequence is the better choice
A calendar sequence is the right choice whenever there is no dated trigger to build on, and that covers more situations than we would like. If you sell to anyone other than ecommerce stores, none of this applies. If your prospects run on a platform other than Shopify or WooCommerce, or sit outside Europe and North America, they are invisible to us whatever they did this week.
We have no LinkedIn signals and no LinkedIn sending. Hiring signals, registry and firmographic triggers, app installs, funding changes, cross-border expansion and review momentum are all marked planned on the signal catalogue, not shipped. Retrieved 22 September 2026.
There is a structural cost no feature removes: age-aware sequencing sends less. Some weeks it enrols almost nobody, because almost nobody did anything. If your team is measured on messages out, that is hard to defend.
The ecommerce leads database describes the pool. Whether the signals fire often enough in your niche is a question your own market answers better than this article can.
Sources
- The buying signal catalogue, Keaz Signals. The "days old, not weeks" position, the "noise dressed up as an opportunity" line, the per-signal store counts and the list of planned-but-not-shipped signals all come from this page. Retrieved 22 September 2026.
- Agent, Keaz Signals. Per-store copy written from the store's own event, and sending pushed to Instantly. Retrieved 22 September 2026.
- Own pool figures: 4.1 million Shopify and WooCommerce stores under watch across Europe and North America, re-scraped weekly. As of September 2026. These figures move and are not permanent.
- No competitor figures, no third-party benchmark studies and no reply-rate claims are cited here. That is deliberate, and the reason is two sections above.
Questions we get
What is a cold email sequence?
A cold email sequence is a scheduled series of messages sent to a prospect who did not ask to hear from you, usually spaced over days or weeks. The scheduling is the part worth arguing about: most sequences space their steps from the day you enrolled someone, which quietly assumes the reason for writing does not age.
How old is too old for a buying signal?
Our published position is that we surface events days old rather than weeks, and that a three-week-old signal is noise dressed up as an opportunity (buying signal catalogue, retrieved 22 September 2026). We are not going to turn that into a single cut-off number for your sequence, because the right window almost certainly differs by signal type and we have not measured it per type.
How many steps should a cold email sequence have?
We do not have a number for you. Every sending vendor publishes one and they disagree, because each is measuring its own customers' offers on its own customers' lists. What matters more than the step count is whether each step still has a current reason behind it.
Can you trigger a sequence the moment a store installs an email tool?
Not to the hour. Stores in our pool are re-scraped weekly, so a signal's age is accurate to roughly a week. Sequences should be designed around that resolution rather than around real-time detection we do not offer.
Does this replace calendar-based sequencing entirely?
No. If your prospects are not Shopify or WooCommerce stores in Europe or North America, or your trigger is not observable from outside the store, there is no dated signal to sequence against and a calendar is the honest fallback.
Builds the signal pipeline behind Keaz Signals. Writes about what the store data actually supports, and what it does not.
Keep reading
Reading about signals is fine. Seeing yours is better.
Access opens per market, in order of signup. When your seat is ready you see which stores in your niche are moving and what we would send them.
Launch my agent