Every team I talk to right now is building a signal-based stack. Job changes. G2 reviews. Web visits. Funding rounds. Hiring spikes. The tools are everywhere and the category is exploding.
But another thing I’m seeing is that most teams are running the same plays off the same signals as every other company chasing their ICP. And then they wonder why the conversion rates fall flat.
I watched this play out with two clients last year. Different companies, same target personas. Both had decent signal coverage. Both were triggering outreach sequences off the same intent data. Both were getting okay-but-not-great results, and neither could figure out why.
TL;DR is that the problem was they’d never stopped to ask, “which signals matter for us, specifically?”
There’s a difference between building a signal stack and building your signal stack. This article is about the second one.
👋 Hi, I’m Kaylee Edmondson. Looped In lands in your inbox every Sunday with one goal: to give you a sharper way to think about demand gen and growth in B2B SaaS. 2k+ marketers are already reading it. If you're not subscribed yet, fix that below.
Why Most Signal Stacks End Up Looking Identical
Everyone starts from the same “menu”.
Intent data platforms hand you a list of signals to track. Your ABM tool has recommended triggers built in. You look at what your competitors are doing and reverse-engineer their plays. Before long, you’ve got a list of 15-20 signals and a rough sense that you should be “acting on” all of them.
But the signals aren’t the edge. The edge is which signals are predictive for your product, your ICP, and your motion specifically.
If you’re showing up in communities asking, “what signals are working for everyone rn?”…you’re asking the wrong question.
Clay wrote a piece last year on what they call GTM alpha - the idea that winning teams use data others don’t have, in plays others can’t run. That framing is right. But most demand gen teams interpret that as “we need more signals” when the actual answer is almost always “we need fewer, better ones.”
Start With Your Best Customers, Not a Signal Menu
The process has to go backwards.
Before you open a signal tool or build a play, you need to understand what was actually true about the accounts that closed, stayed, and expanded. This is the work most teams skip because it feels slow. Or companies are convinced they can’t learn anything from the past. It feels like research instead of execution. But skipping it is what lands you in generic plays that everyone else is also running.
Here’s the exercise I walk clients through:
Pick 5-10 accounts you’d clone if you could. The ones where the deal moved fast, the champion was engaged, the expansion came without you having to chase it.
Write them down.
Then answer these questions for each one:
What was happening at that company in the 60-90 days before they came into your pipeline? Not just “they visited the website.” What was the business context? Were they in a growth phase? Had leadership changed? Were they mid-stack consolidation? Had they just shipped something new?
What did your sales team already know about them before the first call? What research had your AE done that gave them an edge in discovery?
If you had 10 interns researching an account before outreach, what would you have them look for? What information, if you had it, would change your message or your timing?
When you do this across 5-10 accounts, patterns emerge. They almost always do. You’ll start to see 3-5 behavioral or contextual signals that show up consistently in your best accounts. Those become your hypotheses. Everything else is noise until proven otherwise.
Creating Signal Tiering
Once you have your hypotheses, you need a way to organize them. Not all signals carry equal weight for your motion, and treating them like they do is how you end up with a 20-signal stack firing off random acts of marketing.
I use a simple three-tier framework.
Tier 1: High-conviction signals
These are specific, time-sensitive, and directly connected to a problem your product solves. They’re rare, likely for a smaller audience set, but when they fire, they mean something.
An example for a company that sells onboarding tooling: a company just hired its third Customer Success Manager in 60 days. This probably isn’t ironic timing, but is your sign the team is scaling a function that has a real, immediate problem you solve. The signal is specific. The timing matters. There’s a clear message to build around it.
Tier 1 signals are your plays. They’re what you build creative around and automate with care.
Tier 2: Supporting signals
These add context and confirm fit, but they’re not strong enough to trigger a play on their own. A Series B raise plus active SDR hiring is interesting. Layer it on top of a Tier 1 signal and it sharpens your targeting. Use it alone and you’re competing with everyone else who has the same data.
Tier 2 signals belong in your enrichment layer. They help you score and prioritize, but they don’t drive plays.
Tier 3: Noise signals
Every team has a few of these. They felt promising when you added them. You’ve been tracking them for 12-18 months. They haven’t correlated to anything meaningful. But nobody has had the conversation about cutting them because it feels like giving up.
Cut them. The cognitive overhead of managing signals that don’t convert is real, and it crowds out the space you need to think clearly about the ones that do.
Start by mapping everything you’re currently tracking into these three tiers. Most teams will find they’re heavily over-indexed on Tier 3, lightly invested in Tier 1, and confused about where Tier 2 fits.
Building Plays Around Your Tier 1 Signals
When you know which signals are predictive, play design gets cleaner.
A well-built signal-driven play has four components. Before you build anything, you need answers to all four.
The trigger: What exact condition fires the play? “Job change” is not a trigger. “New VP of Revenue Operations hired from a company with $50M+ ARR, into a company currently using a fragmented data stack” is a trigger. The more specific you can get here, the more relevant your outreach will be. Specificity is not over-engineering. Specificity is respect for your prospect’s time.
The context layer: What do you need to know about this account before you reach out? What enrichment should happen automatically before the sequence fires? Think about what a great AE would research before a cold call. Some of that can be automated now. Build it in.
The message: What is the one thing you want to communicate based on this signal? One. If you’re trying to say three things in your first touch, you’re not clear on why the signal matters. Go back and sharpen it.
The timing window: When does this signal stop being relevant? Most signals have a 2-4 week window before the context shifts. A new hire settles in. A compliance event gets handled. If you’re not building timing into your plays, you’re leaving a lot of relevance on the table.
A quick example from a previous client. They sell to mid-market HR teams. Their signal stack was pulling job change data on HR leaders and triggering generic sequences. Response rates were mid at best.
We went back through 12 months of closed-won deals and found that the accounts that moved fastest had one thing in common: the HR leader had been promoted into the role internally, rather than hired externally. Internal promotions meant they were inheriting a tech stack they didn’t choose and were actively evaluating what to keep. Turns out this was a Tier 1 signal. It was specific, time-sensitive, and directly tied to a buying moment.
They rebuilt the play around that one signal. The message became more specific/resonate. The timing changed. Results improved.
Maintaining Signal Hygiene Over Time
Signals decay. Honestly probably faster than we’re even estimating.
A play that worked six months ago may be producing diminishing returns now because the market shifted, because prospects have gotten wise to the trigger, or because three of your competitors started running the same sequence off the same data. This is the reality of signal-based marketing. Nothing stays alpha forever.
Quarterly signal audits, at minimum, are how you can stay ahead of decay. This doesn’t have to be a big process. Here’s what to do:
An easy way to do this is to organize signals to campaigns. Then pull down your campaign data to see which signals are correlating to pipeline and closed won. If a signals has been active with plays running against it for 90+ days and you can’t draw a line to revenue, this has likely become a Tier 3 signal that needs to be demoted or cut entirely.
Review anything you’re tracking but not acting on. Either build something around it or stop tracking it.
Talk to your sales team every quarter about what patterns they’re seeing in discovery. New signals show up there first. An AE who’s done 30 discovery calls in the last 90 days knows things about your buyers that no intent tool can surface.
And treat your signal stack the way you’d treat your tech stack. Regular pruning. Add new things intentionally. Question anything that’s been there a long time without proving itself.
One thing’s for sure, the signal-base marketing noise is only going to get louder. Every team is getting access to more data, better tooling, and faster workflows. The advantage is not having more signals, but in knowing which 2-3 signals are yours to own, and building plays that nobody else can replicate because nobody else did the work to find them.
See ya next week,
Kaylee ✌
And ideally, you’ve already built brand preference before the signal so you aren’t showing up too late.
So often the CMOs I talk to already have a list of vendors or brands in mine when they join a new company.
They’ve either worked with them before or have wanted to work with them for a while.
Signals are no doubt valuable for ABM/outbound motions, but let’s not forget evergreen campaigns that maximize for comprehensive audience reach.