Hlib Storchak · 2026-08-20 · ~9 min readQuick answer
Hiring an outbound agency is the wrong call if you've never run outbound yourself, won't actively manage the engagement, don't have a defined ICP and offer yet, or are shopping on guaranteed meeting counts. A widely cited "only 7% of buyers got outsourced SDRs to work" stat traces back to a 2023 SaaStr post, not new 2026 data, and its real finding was never really about outsourcing itself. It was about who succeeds when they outsource.
A 2023 blog post is still framing your 2026 agency decision
I'm Hlib Storchak. I build and run outbound systems for B2B founders and sales teams, and I've booked 2000+ meetings for B2B clients doing it. I also run a done-for-you outbound practice, which makes me a direct competitor to every agency I'm about to discuss. Keep that in mind as you read this, because I'd rather you know the bias going in than pretend I don't have one.
Here's what got me writing this. If you search anything like "should I outsource my SDRs," you'll land on a wall of 2026-dated guides all repeating some version of the same line: only 7% of companies say outsourced SDRs really worked for them, with 26% saying it sort of worked. It shows up as a header stat on sales-outsourcing landing pages, inside "reality check" teardown posts, and in at least two agency-adjacent blogs using it to argue you should hire them instead of doing it yourself. Almost none of them link to where the number actually came from. I did, and it's from May 2023.
What that SaaStr post actually says, and when it was written
The number traces to a SaaStr post by Jason Lemkin, "Only 7% of You Have Really Gotten Outsourced SDRs to Work," published May 28, 2023, built on an informal SaaStr community poll of 1,200+ founders and operators, not a rigorous, methodology-disclosed research study. Lemkin's own summary: 7% said outsourced SDRs really worked, another 26% said it sort of worked, and the rest, roughly two thirds, saw little or nothing back for the spend. His own take was blunt: "it's hard in practice to outsource something you don't already know well yourself," and he said he'd personally never seen an outsourced team fully replace an in-house one.
None of that is wrong information. It just isn't 2026 data, and it was never a controlled study to begin with, it's one operator's read of an informal poll of his own community. A three-year-old anecdotal post being cited by name-brand-sounding "2026 benchmark" content as if it were fresh is exactly the kind of thing I check before I use a number in front of a client, and it's the same habit I've written about tracing LinkedIn's own recycled quota stat and cold email lift percentages that don't appear on the pages crediting them. Old numbers dressed up as current ones are common enough in this space that checking the publish date has become step one for me, not an afterthought.
Tip. Before you repeat any "X% of teams say outsourcing worked" stat to a client or a boss, click through to the actual source and check the date. If the post doesn't state its methodology or sample beyond "a poll" or "a survey," treat it as one operator's anecdote, useful, but not a benchmark.
The real lesson buried under the headline number
The part of Lemkin's post that actually matters isn't the 7%, it's his explanation for the split. The successful minority weren't lucky, and they didn't necessarily hire a better agency. What set them apart, by his own account, was that they'd already run the function reasonably well in-house first, then stayed heavily involved once it was outsourced: reviewing lists, reviewing scripts, sitting in on calls, treating the outsourced reps as part of the actual team rather than a line item to forget about. The failures mostly hadn't done either of those things. That's not really a stat about outsourcing. It's a stat about buyer readiness, and it holds up regardless of which year you read it in.
That's the frame I actually use with prospects, and it's the one the rest of this article works from. An agency isn't inherently a good or bad bet. It's a good bet for a buyer in a specific state, and a bad one for a buyer in a different state, no matter how good the agency itself is.
Signals 1 and 2: you've never run it, and you won't manage it either
Signal 1: nobody on your team has ever run outbound successfully, even badly. If you're hiring an agency because outbound is a total black box to you, you have no way to judge whether the sequence they're sending is any good, whether the list they built is on-ICP, or whether a 2% reply rate is normal or a red flag. You're not evaluating a vendor at that point, you're trusting one, and trust without the ability to check it is exactly the setup Lemkin's failures shared.
Signal 2: you're hiring the agency specifically so you don't have to think about outbound anymore. That instinct is understandable and it's also the single clearest predictor of a bad outcome. Every agency I'd actually recommend still needs a client-side owner checking in weekly: reviewing reply samples, flagging bad-fit leads, adjusting the offer when something obviously isn't landing. If nobody on your side is doing that, it doesn't matter how good the account manager is. Nobody is watching the thing that's supposed to be watched from both sides.
Signals 3 and 4: no defined offer, and shopping on guaranteed counts
Signal 3: your ICP and offer aren't defined yet. An agency can execute a sequence well. It cannot invent your positioning for you, and most won't try very hard to, because that's genuinely your job, not theirs. If you're hoping the agency will figure out who to target and what to say, you're paying agency rates for work that needs a founder or a product marketer in the room first. I've watched this exact gap sink engagements that had a perfectly competent agency on the other end.
Signal 4: you're choosing based on a guaranteed number of meetings. A guarantee sounds like risk reduction. In practice it usually means a looser definition of "qualified meeting" behind the scenes, since that's the only lever an agency has to hit a number regardless of how your actual market responds that quarter. I've written a full list of these patterns in outbound agency red flags I'd walk away from, and a vague meeting definition paired with a guarantee is close to the top of it.
Signals 5 and 6: a budget too thin to ramp, and wanting a black box
Signal 5: your budget can't survive a real ramp period. Most agencies need 4 to 8 weeks before volume and targeting settle in, and meaningful pipeline usually shows up closer to month two or three, not week one. If a bad first month would end the engagement on cash alone, you're set up to quit right before the version of the program that might have worked.
Signal 6: you actually want a black box, not a partner. Some buyers say they want "full ownership" from the agency, meaning don't loop me into decisions, just send me the numbers. That's the closest thing to a guaranteed failure mode in this list, because it removes the one variable Lemkin's data says matters most: an engaged client-side owner who can tell the difference between a slow start and a program that's quietly going nowhere.
Agency-ready vs not ready, side by side
| Signal | Agency-ready | Not ready yet |
|---|---|---|
| In-house outbound experience | Someone on your team has run it, even a rough version | Nobody has ever sent a cold sequence |
| Weekly involvement | A named owner reviews replies and lists weekly | Intent is to "hand it off" entirely |
| ICP and offer | Defined, tested, written down | "The agency will figure out who to target" |
| How you're buying | Judging on process, reporting, and fit | Judging mainly on a guaranteed meeting count |
| Budget horizon | Funded for 3+ months before judging results | One slow month would end it |
| What you want from them | A partner you check in with | A black box you never touch |
When hiring an agency is genuinely the right call
None of this means agencies are a bad bet in general, and I'd be misrepresenting my own client base if I said otherwise. An agency is the right call when you already know roughly what good outbound looks like for your business, you have a named person who'll spend an hour or two a week actually reviewing the work, your ICP and offer are settled enough to brief clearly, and you're funded to judge results at month three rather than week three. Under those conditions, an agency mainly buys you speed and headcount flexibility on a system you're already equipped to evaluate, which is a genuinely good trade. The failure mode isn't the agency model itself. It's using it to skip the parts of the work that were never actually the agency's job.
What getting this wrong actually costs
It helps to put a number on the downside, built on assumptions you should swap for your own. Say you sign a mid-market retainer at $6,000/month (inside the $4,000 to $8,000/month range LeadRiver's 2026 benchmark clusters around for this tier), with a $5,000 setup fee and a 3-month minimum term, a common structure per Salespipe's 2026 pricing guide. If none of the six signals above are in place, you're statistically closer to the roughly two-thirds of buyers in Lemkin's informal poll who saw little back, so treat the realistic downside as close to the full spend: 3 months x $6,000 plus the $5,000 setup, or about $23,000, not counting your own team's time briefing and chasing the agency along the way. Compare that to the cost of getting agency-ready first: a few weeks of a founder or a fractional GTM lead defining ICP and offer, then testing a small in-house batch before you ever sign a contract, which usually costs a fraction of that $23,000 and tells you, cheaply, whether you're actually ready to hand it off. Run your own numbers against your own retainer quote and minimum term before you sign anything.
What I actually tell prospects who ask me this
The mistake I see most often, across done-for-you clients, fractional engagements, and teams I help stand up their own in-house function, is the same one Lemkin flagged in 2023: wanting outbound handled by someone else specifically because it feels too unfamiliar to supervise. When a prospect tells me that outright, I usually tell them the honest thing even when it costs me the deal: start smaller, define the offer first, stay in the room for the first month, whichever way you go. An agency, a fractional lead, or your own hire all fail the same way when nobody on the buying side can tell good work from bad. That's the actual thing worth fixing before you pick a vendor.
Key takeaways
- The widely repeated "only 7% of buyers got outsourced SDRs to work" stat traces to a May 2023 SaaStr post by Jason Lemkin, an informal poll of 1,200+ founders, not new 2026 research.
- The post's real finding wasn't about outsourcing being risky. It was that success tracked with prior in-house experience and heavy ongoing client-side involvement, not with which agency got hired.
- Six signals predict a bad outcome: no in-house experience, no plan to stay involved, an undefined ICP and offer, buying on a guaranteed meeting count, a budget too thin to survive a ramp, and wanting a black box instead of a partner.
- None of that makes agencies a bad model. It makes them a bad fit for a buyer who isn't ready to evaluate the work.
- A cost model built on stated assumptions (a $6,000/month retainer, 3-month minimum, $5,000 setup) puts the realistic downside of a mismatched engagement around $23,000, which is worth comparing to the cost of getting ready first.
FAQ
Is it true that only 7% of companies succeed with outsourced SDRs?
That figure comes from an informal SaaStr community poll published in May 2023 by Jason Lemkin, not a controlled 2026 study. It's a real, attributable data point from over 1,200 respondents, but it's an operator's anecdotal read, not a rigorous benchmark, and it's three years old despite showing up in content dated 2026.
What actually separated the successful 7% from the rest?
By Lemkin's own account, the successful minority had usually run the function reasonably well in-house first and stayed closely involved after outsourcing it, reviewing lists, scripts, and calls. Most of the unsuccessful group had done neither.
How do I know if my company is ready to hire an outbound agency?
Check for six signals: someone on your team has run outbound before, you'll assign a named owner to review the work weekly, your ICP and offer are already defined, you're evaluating on process rather than a guaranteed meeting count, your budget can survive a 2 to 3 month ramp, and you actually want a partner rather than a hands-off black box.
What should I do instead if I'm not ready for an agency yet?
Spend a few weeks defining your ICP and offer and running a small in-house test batch, alone or with a fractional GTM lead, before signing an agency contract. It's usually far cheaper than a mismatched retainer and tells you plainly whether you're ready to judge an agency's work once you do hire one.
Does this mean outbound agencies don't work?
No. Plenty of my own clients run agencies successfully, and I still recommend the model when the fit is right. The failure mode isn't the agency, it's using one to skip the ICP, offer, and ongoing-involvement work that was always the buyer's job, not the vendor's.