Full disclosure, because trust is the whole point.
I take no money from any tool in this issue. No sponsors, no affiliate links, no vendor paid to be here and none of them saw this before you did. Every tool is scored against the same published rubric, and I name the ones that come up short by name. Read me skeptically, which is the point.
The category that sells a number nobody measures
Sales engagement is the sending layer. It is the software that takes a list and turns it into email landing in somebody's inbox, with follow-ups, call tasks and a reply-handling queue bolted on. It is one of the oldest categories in GTM software and it is currently being squeezed from both directions at once.
From below, cheap AI-native senders have collapsed the price. Smartlead starts at $39 a month with unlimited mailboxes. Instantly starts at $47. Saleshandy starts at $34 billed annually. From above, the data platforms have absorbed the sending function entirely: Clay now ships a Sequencer, Apollo bundles sending with its database, and Unify sells deliverability as a named product with managed mailboxes at $25 each. The incumbents in the middle, the ones who built this category, are responding by renaming themselves. Outreach rebranded from outreach.io to outreach.ai in April 2026 and now calls itself an agentic revenue platform. Salesloft merged with Clari in December 2025 and no longer publishes a price at all.
Here is the number that reframes everything.
Smartlead publishes a benchmark hub built from more than 850 million emails sent through its own platform. On that data, the median cold email reply rate is 0.74%. That is one reply for every 135 contacts. The top 5% of senders reach 3.85%. Smartlead also publishes the median bounce rate at 1.54%, and it refuses to publish an open rate at all, on the grounds that open rate is no longer a reliable measure for cold email. I went looking for a vendor with an incentive to flatter these numbers and instead found the one that publishes the harshest version of them.
Now hold that against what the category sells. Every tool in this issue is bought on the promise of deliverability. Inbox placement is the entire value proposition of the volume tier and a headline feature of the rest. So I went looking for the number.
How I scored it
Five pillars, 100 points, retuned for this category because deliverability is the axis where vendor claims are actually falsifiable.
AI-native vs. bolted-on (30). Built as an AI-first sending layer, or sequencing grafted onto a CRM you already pay for?
Deliverability and sender safety (25). Inbox rotation, warmup, throttling, domain health, and crucially whether the vendor publishes anything measurable about its own performance.
Actionability (20). Does it drive the next step, or just report what happened?
Integration and workflow fit (15). Does it sit in the stack you have, or demand you rebuild around it?
Trust and governance (10). Security posture, incident history, pricing transparency, and whether a buyer can size the vendor at all.
Scores below are my framework-based read of each tool's public positioning and category archetype, not lab-tested benchmarks. A map, not gospel.
The archetypes (this is the real insight)
Strip away the feature lists and all 21 tools fall into five buckets, plus one genuinely new wrinkle.
The Command Centers. Outreach, Salesloft, Amplemarket. Enterprise suites where sequencing is one module inside a revenue-orchestration story. None of the three publishes a price. Outreach lists four packages sized in AI credits and every tier reads "Request pricing." Salesloft's pricing page is a booking form with no plan names and no numbers on it.
The Volume Engines. Smartlead, Woodpecker, Instantly, Saleshandy, lemlist, Reply.io, Klenty, Outplay. Priced on emails sent rather than seats, with unlimited mailboxes, rotation and warmup built in as table stakes. This is where the real sending happens and where the published benchmarks live.
The Data Layers. Clay, Apollo, Unify, ZoomInfo. Platforms that own the list and have moved downstream to own the send. Clay raised $115M at a $7.1B valuation in September 2026 and shipped a sequencer. Apollo acquired Pocus in March 2026. This is the squeeze from above and it is the structural story of the category.
The Platform Bolt-Ons. HubSpot Sales Hub, Salesforce Sales Engagement, Yesware. Sequencing attached to a system of record. One mailbox per rep, hard daily caps, and no rotation or warmup documented anywhere.
The Agents. Artisan, 11x. The agent does the sending. You configure an ICP and a knowledge base and it runs.
The wrinkle: The Identity Layer. La Growth Machine prices per sending identity rather than per seat or per send, and gives each identity its own dedicated mobile proxy. It is the only tool here that treats the human sender, not the software seat and not the email volume, as the unit being protected.
Knowing which bucket a tool is in tells you more than any feature comparison will.
The scorecard

Verdict band: 90+ buy with confidence, 75 to 89 strong with caveats, 60 to 74 situational, under 60 skip or wait. (Framework-based read of public positioning, not lab-tested benchmarks.)
What actually separates the winners
The category grades its own homework and there is no referee. This is the villain, and it is structural rather than any one vendor behaving badly. Every deliverability tool a buyer can actually run, Google Postmaster Tools, Validity Sender Score, GlockApps, MailReach, measures the sender's own domain and IP reputation. None of them measures the platform. So the only party in a position to publish platform-level inbox placement is the platform, and almost nobody does. Unify published a five-way deliverability comparison of Outreach, Salesloft, Apollo, Instantly and itself in October 2026, and its own FAQ on that page answers the question "Which platform has the best inbox placement?" with the sentence "This documentation review cannot establish a winner." The same page states "It is not a controlled inbox-placement benchmark." A vendor published a competitive comparison and then printed the disclaimer that it proves nothing. That is the most honest document in the category and it is an admission of the hole.
Two vendors publish an inbox placement rate and they are not measuring the same thing. Saleshandy published a study in June 2026 stating that across 53 million cold emails sent through its platform, the average inbox placement rate was 95.2%, with 1.8% landing in spam. Smartlead publishes roughly 99% inbox placement, 99.04% on Outlook and 98.83% on Gmail, built from its 850 million email sample. Those look like the same metric and they are not. Smartlead's figure covers warmup emails, which are messages its network sends to other mailboxes inside its own network, and Smartlead says so on the page, calling it a health check rather than a target. Saleshandy's figure covers real campaign sending, which is the number buyers think they are getting, and it comes with no seed-list design and no explanation of how placement was measured, so it cannot be reproduced. The honest summary is that the category's two published placement numbers are one self-administered health check and one unreproducible claim.
The only third-party number in the category was bought by the vendor it flatters. Apollo puts 94% on its deliverability product page, labelled "Independently benchmarked by the Tolly Group." The underlying document is Tolly Report 226102, dated April 8, 2026, and Tolly names Apollo as its sponsor on the report. The test was a live campaign of 384 prospects across 205 companies, capped at 50 emails a day, run without the usual four to six week warming period. 361 of 384 emails were accepted by receiving servers. That is server acceptance, not inbox placement, and the number of sending domains is not stated. It is a real test, properly documented, and it is the best evidence anyone in this category offers. It is also 384 emails, paid for by Apollo, run by a lab piloting its own new service. That gap between "best evidence available" and "evidence" is the whole problem, and it cost Apollo the top slot.
The bottom line
If you are sending real cold volume, buy a Volume Engine and treat the published benchmarks as the tiebreaker. Smartlead at 82 takes it, not because 99% warmup placement means much, but because it is the only vendor that publishes its own median reply rate at 0.74% and its own median bounce at 1.54% with the sample size and the inclusion rule attached. A vendor willing to print a number that unflattering is telling you something about the rest of its numbers. Woodpecker at 77 is the methodology winner outright, with per-table sample counts across 56,614 campaigns and a stated exclusion policy, and it is also the only vendor here whose financials you can check: revenue fell from roughly 26.5M PLN in 2023 to 22.1M in 2024 to 18.8M in 2025, with a net loss last year and headcount cut from 73 to 54. Buy the benchmarks, price in the restructuring.
If you already own the data layer, Clay at 81 and Apollo at 79 are the pragmatic picks, with the caveat that Clay publishes no measured deliverability data at all for the sequencer it is asking you to send from.
If you are running follow-up rather than cold outbound, the bolt-ons are fine and cheaper than they look. HubSpot at 75 caps you at 500 sends per user per day on Professional, which costs $90 per seat per month plus a $1,500 onboarding fee, and that cap is also drawn against your Gmail or Microsoft ceiling, so the effective number is the lower of the two. That is a real limit and it is the correct limit for inbound follow-up.
Skip the agents for now. Artisan at 59 publishes no price and its own help documentation contradicts its marketing page about whether the agent or the human handles replies. 11x at 53 lists three different starting prices on one pricing page, $2,000, $3,000 and $3,750 a month, and the 2025 reporting about its customer logos and its ARR accounting remains unresolved in public. Yesware at 50 is the bottom, and not because it is bad at what it does: it sends from your own inbox with no rotation and no warmup by design, its own scale figures on its own site are still the 2022 numbers from before Vendasta acquired it, and the free tier caps you at ten campaign recipients a month.
The bigger pattern. Issue #007 took apart data enrichment and found the same hole in the same place: a category selling accuracy where almost nobody published a measured accuracy rate, and where the one number everyone quoted traced back to a vendor with an interest in it. Sending has the identical shape. The enrichment layer could not prove the data was right and the sending layer cannot prove the email arrived. Two issues, two categories, one missing referee. The pattern is not that these vendors are dishonest. It is that nobody has built the independent measurement layer that would let them compete on proof, so they compete on features instead, and buyers pay for features that may or may not do the job.
Next issue: Intent Data and Buying Signals
The category that claims to know who is in market before they tell you. I teased it at the end of #008 and it is next, properly this time. Same question, pointed at a harder target: when a vendor says an account is showing intent, what exactly did it observe, and can you check it?
GTM Tool Time is an independent review. No vendor pays to be here. Reply and tell me which category to tear down next.
Marty
Marty Berman is the publisher of GTM Tool Time and Founder of Scalient IQ.
See the full Sales Engagement directory and the scoring rubric at gtmtooltime.beehiiv.com.