In January, our growth team booked 188 meetings and generated $3.3M in pipeline. Two people ran the whole thing.

I've gotten more questions about this than anything else I've posted in the past year, so I want to explain it properly. Not as a hype post, but as an honest account of what we actually built and how we think about it.

The short version: we stopped doing cold outbound entirely. Everything we send is triggered by a signal. That one change made everything else easier.

How we used to think about outbound

When I was at Ramp, outbound meant building a list, writing sequences, and grinding through volume. You'd optimize the message, A/B test subject lines, adjust cadence. The entire effort focused on making the email better.

It worked. We had good results. But there was a ceiling on what optimization could do, and we kept running into it.

The problem wasn't the email. The problem was that we were still guessing at who wanted to hear from us and when. We'd filter by company size and title and industry, which is fine, but none of that tells you whether this specific person is actively thinking about this specific problem right now.

When I started Unify, I wanted to see what happened if we solved the timing problem instead of the message problem.

What a signal actually is

A buying signal is something a person or company does that suggests they're thinking about a problem you solve. It's not demographic data. It's behavioral data. Here are the ones we use most:

Job change. A VP of Sales or Head of Growth just joined a company in your ICP. New leaders almost always evaluate their current tools and processes in the first 90 days. They're not locked into what was already there. They're forming opinions. That's a real window.

Funding event. A company just raised a Series A or B. They're about to scale. They need infrastructure to support that growth and probably don't have it yet. The timing is almost always right.

Pricing page visit. Someone at a target account visited your pricing page and didn't convert. They were interested enough to look. Something stopped them. A message arriving within 24 hours closes a lot of those loops.

Competitor engagement. Someone at a target account liked a competitor's post or commented on their founder's content. They're actively thinking about the category. You're not interrupting. You're showing up in context.

Champion movement. Someone you've talked to before just started a new role at a company you want to work with. They already know what you do. They're in a new seat. Easiest call in the book.

The play anatomy

Every outbound play we run follows the same structure: a signal triggers a watch condition, the watch condition matches against our ICP, a task gets created for the rep or an automated sequence fires, and the message references the signal directly, not in a creepy way, just in a way that makes clear this isn't random.

Something like: "Saw you just joined [Company], congrats. We've helped a few other GTM leaders there think through their outbound setup in the first 90 days. Worth a quick call?" That's it. The signal does most of the work. You're not cold. You're responding to something they already did.

What this changes about your metrics

When you run signal-based outbound, your reply rates look nothing like industry averages, because you're not doing the same thing. Our response rates on job change plays are consistently 3-5x higher than what we see on unsegmented sequences. Not because the writing is better. Because the timing is right.

The other thing that changes is rep morale. Our growth team doesn't feel like they're spamming people. They're running targeted plays based on real context. That's a different mental model, and it shows in how they approach the work.

Where to start

Most teams that try to adopt signal-based outbound get stuck because they try to build everything at once. They want the full stack before they run a single play.

Start with job changes. It's the cleanest signal, the most accessible data, and the easiest play to run manually before automating anything. Find 20 companies you want to work with. Set up alerts for when a VP-level hire joins their GTM team. Send that person a thoughtful message within 72 hours.

Do that for 30 days. You'll never go back to cold lists. That's what we built. I'll keep sharing more of it as we go.

In October 2024, we announced Unify's $12M Series A. By the end of the month we had booked 494 meetings and generated $6.6M in pipeline from the announcement alone.

Most founders treat a fundraise announcement as a PR moment. It's actually the most powerful outbound play you'll ever get to run, if you treat it that way and move fast enough. Here's what we did.

The 48-hour window

When a company you've been following announces a raise, there's a period of 48 to 72 hours where everyone in their network is paying attention. LinkedIn engagement spikes. Email response rates spike. People who would never reply to a cold note will respond to a congratulations message.

That window is real. Most companies let it close without doing anything with it.

The preparation (two weeks before going live)

We started building about two weeks before we announced. Three things in parallel.

The list. We built an outbound list of every company that fit our ICP and segmented it three ways: warm contacts, cold ICP match, and customers of companies we compete with. Each segment would get different messaging.

The sequences. We wrote every email and LinkedIn message before we hit publish on the announcement. Not rough drafts. Final versions. We wanted to be sending within an hour of going live, not spending two days writing copy while the window was already closing.

The content calendar. We planned 10 to 14 days of LinkedIn content that would run after the announcement. Not all of it was about the raise. Some of it was product content, customer stories, team hires. The raise was the attention-getter. The follow-up content kept people in the feed.

The announcement itself

We published at 8am ET on a Tuesday. By noon the post had more engagement than anything we'd put out before. The reason wasn't luck. We'd spent two weeks quietly seeding it. Investors, advisors, customers, team members all knew it was coming and had agreed to engage with it the moment it went live.

That coordination looks organic from the outside. It's not. It's deliberate. You're not gaming the algorithm. You're giving your network the context to show up on purpose.

The three outbound plays

Within the first 24 hours, three plays ran simultaneously.

The warm list. Every person we'd already talked to and not closed got a personal note that referenced something specific from our last conversation and mentioned we'd just closed our Series A. These had the highest conversion rate of anything we ran.

The ICP cold list. Every company that matched our ICP got a short direct message. We referenced the raise not as a flex but as context: we just closed our Series A and are building out our first class of design partners. If you're thinking about outbound in the next 6 months, worth a conversation. Short. Easy to say yes to.

The competitor customer list. Companies actively using tools we compete with got a message that acknowledged what they're using, explained one specific thing we do differently, and asked for a 20-minute comparison call. Response rate was lower but meeting quality was high.

The content loop

In the 10 days following the announcement we published: a behind-the-scenes post on what we'd built to get to the raise, a customer case study, a product update showing what the raise would fund, and a team hire post. Each had its own engagement. People who missed the announcement saw it through the follow-up content. The announcement was a two-week campaign that kept re-surfacing, not a single moment.

Applying this outside of fundraising

You don't need to be raising a Series A to run a version of this. Any high-signal company moment creates the same window: a major product launch, a named customer you can announce publicly, a hire with their own following, a partnership your audience will recognize.

The frame is the same: something significant just happened, and that's a genuine reason to start a conversation that didn't exist yesterday. Build the list before, write the sequences before, plan the content calendar before. The window is short. Don't build while it's open.

If you're planning a fundraise or major launch in the next few months and want to talk through the playbook, my DMs are open.

Over the past year I've talked to probably 50 companies that used AI SDR tools. Almost every single one has churned or is planning to. This isn't a hot take. It's just what's happening.

The interesting part isn't that AI SDRs failed. It's why they failed. Because the reason is fixable, and understanding it points directly at what actually works.

What went wrong

AI SDR tools are genuinely good at one thing: generating personalized emails at scale. The personalization is real. The copy quality is solid. The volume is impressive. The problem is that personalization without targeting is still cold outreach. It just feels more personal when it's cold.

When I talk to churned customers, they tell the same story. They built a list based on ICP criteria (company size, title, industry) and started sending. Response rates were okay in month one. By month three they were down. By month six they were back to the same level as their old generic outreach, except now they'd sent a lot more of it and it felt more invasive.

The tool was doing exactly what it was supposed to do. It was sending personalized messages. It just had no idea whether any of those people had any reason to care right now.

The targeting problem

Cold outbound thinks about targeting like this: who fits the profile of someone who might buy from us? That's a demographic question. It produces a list of people who match certain attributes: title, company, industry, headcount. Not a bad starting point. But it misses the most important variable: timing.

The same person who would never reply to an email in January might be actively looking for a solution like yours in March. Not because they changed their title. Because something happened. A budget got approved. A new leader joined. A competitor deal fell through. A trigger event created a window that didn't exist three months ago. Demographic targeting can't see trigger events. It just sees attributes that haven't changed.

What signal-based outbound actually is

Signal-based outbound inverts the targeting logic. Instead of starting with "who fits the profile," you start with "what just happened, and who does that touch?" Job changes. Funding events. Pricing page visits. Competitor engagement. Champion movement. You identify the signal, you identify who it touched, and you send a message that references the signal as the reason you're reaching out. The message doesn't have to be clever. It just has to arrive at the right moment.

What the data shows

At Unify, our response rates on signal-triggered plays run consistently 3-5x higher than industry averages on standard cold outbound. We're not exceptional at writing emails. We're just sending them at the right time. The more important metric is pipeline quality. Meetings that come from a genuine trigger event convert at a higher rate through the full sales cycle because the prospect already had context.

What the AI SDR role actually becomes

The AI SDR as a blunt-force personalization tool is dying. But AI-assisted outbound, where the system identifies signals and the human crafts context-aware messages based on those signals, is going to be everywhere. The role isn't going away. It's restructuring.

The people who figure out how to run signal-based plays, who understand the triggering logic and can diagnose why a play isn't converting, who can build and iterate on a playbook, those people are going to be extraordinarily valuable. I've been calling them growth engineers for about 18 months. The demand is already there. The supply hasn't caught up yet.

If you're evaluating AI outbound tools right now: ignore the demo email. Ask them how they decide when to send it. That question tells you everything.