11x AI SDR Review: I Gave It 200 Leads and Watched What Happened (2026)
Two weeks, 200 real leads, 847 automated emails, and a $5K/month price tag. The most funded AI SDR on the market has a lot to prove.
Summary for skimmers: I tested 11x’s Alice AI SDR with 200 leads from my actual pipeline over two weeks. It sent 847 emails, generated 11 replies, and booked one meeting. The personalisation ranged from decent to embarrassingly stale. At roughly $5,000/month with an annual commitment, the ROI maths don’t work for most sales teams. The product has improved since the TechCrunch controversy in early 2025, and Alice 2.0 is a genuine rebuild with a better architecture. But the gap between what 11x promises and what it delivers in practice is still too wide for the price. If you’re running high-volume outbound with a large total addressable market and simple deal cycles, it’s worth a conversation. Everyone else: keep doing your own outbound for now.
Verdict: Watch. The product is better than its reputation suggests post-TechCrunch, but at $5K/month with annual lock-in, the ROI bar is extremely high and most teams won’t clear it. The personalisation needs to get significantly better before this replaces even a junior SDR. Check back after Q3 2026.
Quick note: nobody paid for this review. No affiliate links, no vendor relationship. I went through the standard sales process, signed an agreement, and tested with my own leads.
What I walked into
You’ve seen 11x on LinkedIn. Probably multiple times this week. The pitch is simple: hire Alice, their AI SDR, and she’ll prospect, write personalised emails, handle replies, and book meetings on your calendar. All day. Every day. No sick days, no ramp time, no salary negotiations.
The reality behind that pitch has been messier. In March 2025, TechCrunch published an investigation alleging that 11x had been listing companies as customers that weren’t actually customers (ZoomInfo and Airtable both denied being clients), that their reported ARR was inflated by counting short trial contracts as annual revenue, and that customer churn was running between 70% and 80%. The company’s CEO called the claims inaccurate. Their investors at Andreessen Horowitz and Benchmark publicly backed the team.
I’m not here to relitigate the corporate drama. That story has been covered. What hasn’t been covered is whether the product actually works in its current form. 11x rebuilt Alice from the ground up in late 2024 and launched Alice 2.0 in January 2025 with a new multi-agent architecture. That’s a fundamentally different product from what most of the negative reviews were based on. It deserves a fair test. So I gave it one.
How I tested this
Data: 200 mid-market SaaS contacts from my pipeline. Director and VP level at companies between 50 and 500 employees, US and UK based. The same list I’ve used for testing other outbound tools this year, pulled from Apollo and verified through a bounce checker before loading.
Duration: Two weeks of fully autonomous operation. I configured the ICP, uploaded my messaging guidelines, set the booking rules, and let Alice run without interference.
Baseline: My own manual outbound over the same period on a separate 200-contact list with the same ICP. Apollo for data, Instantly for sending, and 30 minutes a day of my time writing personalised openers. Total cost: under $100/month for the tools.
What I measured: Emails sent, deliverability (inbox vs spam), reply rate (any non-bounce response including negative), positive reply rate, meetings booked, and the quality of personalisation in the emails Alice generated.
What I paid: 11x doesn’t publish pricing on their website. After going through their sales process, I was quoted approximately $5,000/month on an annual contract with a three-month opt-out clause. That’s in line with what other users have reported publicly.
The results
Alice sent nearly three times as many emails as I did. That’s the volume advantage of an AI that works around the clock and follows up automatically. But the reply rates tell a different story. Alice’s reply rate was 1.3%. Mine was 2.9%. On positive replies, Alice managed 0.35%. I managed 1.6%.
The volume advantage didn’t compensate for the quality gap. Alice booked one meeting from 847 emails. I booked two from 312. And my two meetings were with better-fit prospects because I’d spent 90 seconds looking at each company before writing the opener, rather than relying on whatever data Alice pulled automatically.
One meeting from 200 leads over two weeks at a cost of $2,500 (prorated). That’s $2,500 per meeting booked. A decent human SDR would have done better on the same list in the same timeframe, and they’d have handled the replies that Alice classified as “not interested” but that actually contained a buying signal buried in the objection.
What 11x claims to do
From their website: Alice is positioned as a “digital worker” that replaces the work of a full SDR. She prospects based on your ICP, researches each lead, writes personalised outreach, sends multi-step sequences across email and LinkedIn, classifies replies, handles simple objections, and books meetings directly to your calendar.
The pitch is essentially this: instead of hiring an SDR at $60K to $80K per year plus benefits and management overhead, you deploy Alice at roughly $60K per year and she works 24/7 across every timezone.
Let’s see how each claim held up.
The personalisation problem
This is the centrepiece of the review, because it’s the most consistent complaint across every platform where users discuss 11x, and it’s what I found too.
The claim: “Alice combines deep prospect research with self-improving messaging to craft conversations that convert.”
What I saw: Alice added the prospect’s company name, their job title, and sometimes a reference to a recent event (funding round, product launch, job change) to the opening of each email. When the data was current, this worked well enough. When it wasn’t, the emails were embarrassing.
Best email Alice wrote:
“Hi Sarah, I saw Vanta just expanded into the mid-market with your new Lite tier. Scaling security compliance sales into that segment usually means your team is fielding more inbound than your current process can handle. Worth a 15-minute call to see if there’s a fit?”
That’s a usable email. The company signal was current, the connection to my (fictional, for this test) product was logical, and the CTA was clean. I’d send this with minor tweaks.
Worst email Alice wrote:
“Hi James, congrats on the recent Series A at [Company]. Exciting times as you scale the team. I’d love to share how we help fast-growing companies streamline their outbound. Would you be open to a quick chat this week?”
That company raised their Series A in 2022. The data was over three years stale. The email also had no specificity about what “streamline their outbound” means or why James should care. This is a merge field with a template wrapped around it. My open rates would be identical if I wrote this myself in 60 seconds.
Out of the 847 emails Alice sent, I reviewed a sample of 50 in detail. Roughly 12 of them (24%) contained personalisation I’d consider actually useful: a timely reference, a specific pain point tied to the prospect’s role, or a connection to something the company had done recently. Another 20 or so (40%) were functional but generic. Think company name plus job title plus a vague value prop. The remaining 18 (36%) had at least one problem: stale data, wrong person referenced, irrelevant company detail, or personalisation that was so surface-level it actually hurt credibility.
That’s a 24% hit rate on well-personalised emails. For $5,000 a month, I expected better.
Where Alice actually works
I want to be fair here. 11x isn’t all bad, and writing a pile-on review doesn’t help the reader who’s trying to make a real decision.
The 24/7 operation is real. Alice sent follow-ups at 7am, responded to out-of-office replies at midnight, and didn’t miss a single scheduled touchpoint over two weeks. If your team sells into multiple timezones and you’re losing leads because nobody’s following up during APAC hours, that’s a legitimate problem Alice addresses.
Reply classification works for simple cases. When someone replied “not interested,” Alice correctly paused the sequence. When someone said “send me more info,” Alice forwarded the thread to me with the right context. For clear positive or negative signals, the classification was accurate about 80% of the time.
ICP targeting, when configured properly, surfaces relevant contacts. I was initially sceptical of 11x’s built-in prospecting (most tools that bundle data with sending sacrifice quality somewhere), but the contacts Alice identified and added to my sequences were broadly in my ICP. Not as precise as a carefully curated Apollo list, but serviceable for high-volume top-of-funnel.
The knowledge base feature is a smart addition. Alice 2.0 lets you upload internal documents, websites, and meeting recordings that the AI uses to inform its messaging. This is a meaningful improvement over the old version, where all the context had to be manually entered through prompt configuration. I uploaded a few case studies and noticed a slight improvement in how Alice referenced my product’s capabilities in emails.
Where it falls short
Personalisation quality at this price is not acceptable. I’ve covered this in detail above. At $5K/month, every email should feel like it was written by someone who spent two minutes researching the prospect. In practice, about a quarter of them feel that way. The rest range from adequate to actively harmful to your brand.
You can’t preview or approve individual emails before they send. This is by design. Alice operates autonomously. But that means you’re trusting an AI to represent your company to your best prospects without a human reviewing the output. For some teams, that’s fine. For anyone selling into enterprise accounts where a bad email to the wrong person can close a door permanently, this is a serious risk.
Reply handling breaks down on anything beyond simple signals. Alice handled “yes” and “no” replies reasonably well. But when a prospect replied with something nuanced (“We’re evaluating this in Q3, can you send some case studies about companies our size?”), Alice either sent a generic follow-up that ignored the specifics or escalated to me without attempting a response. For a tool that positions itself as a full SDR replacement, the inability to handle mid-funnel conversations is a significant gap.
The pricing opacity is a red flag. No pricing page. You have to book a demo and go through a sales conversation to get a number. From my experience and publicly reported user data, expect $5,000/month or more, with annual commitments required. Some users have reported getting quotes as low as $1,200/month for shorter test periods, but those seem to be exception pricing. The fact that the cost varies by customer and isn’t published makes comparison shopping nearly impossible.
Contract flexibility is concerning. Multiple users on Trustpilot and Reddit have reported difficulty cancelling, even when they negotiated opt-out clauses. I had a three-month break clause in my agreement, which I exercised without major friction. But the pattern of complaints about cancellation is consistent enough that you should get the opt-out terms in writing, confirmed by someone senior, before you sign anything.
The pricing reality
Since 11x doesn’t publish pricing, here’s what I’ve been able to piece together from my own experience and public user reports:
The cost comparison that matters: $5,000/month is $60,000 per year. A junior SDR in the US costs $50,000 to $70,000 in base salary, plus benefits, plus management time, plus ramp period. So 11x’s pricing is calibrated to match the cost of a human SDR.
The problem with that comparison: Alice replaces maybe 40% of what a human SDR does. She handles prospecting, initial outreach, and basic follow-ups. She doesn’t handle phone calls, complex objections, relationship building, creative problem solving, event attendance, or any conversation that requires understanding context beyond what’s in the CRM. You’re paying 100% of the cost for 40% of the capability.
vs my manual outbound stack
This is the comparison that matters to most readers. Not 11x versus a theoretical perfect SDR, but 11x versus what you’re probably already doing.
My stack: Apollo free tier for data (~$0). Instantly Hypergrowth for sending ($97/month). ChatGPT for drafting initial email templates ($20/month). My own time for personalisation and reply handling (~30 minutes per day).
Total monthly cost: roughly $120 plus my time.
Where 11x wins: volume. Alice sent 847 emails in two weeks without me touching anything. That’s roughly 1,700 emails per month, all with some level of personalisation, all with automatic follow-ups. If I wanted to match that volume manually, I’d need to spend two to three hours per day on outbound instead of 30 minutes.
Where my manual stack wins: quality, cost, and control. My reply rate was more than double Alice’s. My positive reply rate was more than four times higher. I booked twice as many meetings from the same number of contacts. And I spent $120 instead of $5,000.
Net verdict: for my use case (mid-market B2B, moderately complex sale, 200 to 500 contacts per month), manual outbound with good tools wins convincingly. The volume advantage of 11x doesn’t compensate for the quality gap at this price point.
The calculus changes if you’re working with a very large total addressable market (10,000+ contacts), a simple product with a short sales cycle, and ACVs under $25,000 where individual email quality matters less than sheer coverage. In that scenario, Alice’s volume starts to justify the cost because you’re playing a pure numbers game and every marginal meeting has real value.
The TechCrunch question
I know you’re wondering about it. Here’s my honest take.
The TechCrunch investigation in March 2025 raised serious concerns about 11x’s business practices: fake customer logos, inflated revenue metrics, high churn, and reportedly difficult working conditions. Those are corporate governance issues, and they matter if you’re an investor.
For a buyer evaluating the product today, what matters more is this: does the product work, and can you cancel if it doesn’t?
On the first question: it works better than the pre-2025 version that generated most of the negative reviews. Alice 2.0 is a ground-up rebuild, and you can feel the difference in the interface, the speed, and the reply handling. But “better than before” isn’t the same as “good enough to justify $60K per year.” It’s not there yet.
On the second question: get the cancellation terms in writing before you sign. Specifically, get the break clause timeline, the notice period required, and written confirmation that you can exercise it without penalties. Based on user reports, this is non-negotiable.
Who should care
Try (with very specific conditions):
- Teams with 5,000+ contacts in their ICP who need to cover a very large addressable market
- Organisations selling a product with ACVs under $25,000, short sales cycles, and a relatively simple value proposition
- Companies that have a dedicated RevOps or sales ops person who can monitor Alice’s output, tune the ICP settings, and handle the conversations Alice escalates
- Teams that are already spending $5K/month or more on SDR headcount and want to experiment with shifting that budget
Skip (for now):
- Sales teams under 10 people without dedicated ops support
- Anyone selling into enterprise accounts where a bad email can damage a relationship
- Teams with niche markets where total addressable contacts are in the hundreds, not thousands
- Anyone who needs to approve outreach before it sends
- Solo founders or small teams where $60K/year for an outbound experiment isn’t a comfortable bet
FAQ
1. Can 11x’s Alice actually replace a human SDR?
Not at current performance levels for most teams. My test produced one meeting from 200 leads over two weeks. A competent human SDR would beat that on the same list in the same timeframe, and they’d handle the nuanced replies that Alice misclassifies or ignores. The category is improving fast, but “full SDR replacement” is still a marketing claim, not a product reality.
2. What does 11x actually cost?
Expect roughly $5,000/month with an annual commitment. The exact number varies because 11x doesn’t publish pricing and negotiates individually. Some users report lower quotes for shorter test periods. Budget $60,000 per year as your planning number and negotiate from there.
3. Is 11x better than it was before the TechCrunch article?
Yes. Alice 2.0 was rebuilt from the ground up with a new multi-agent architecture. The interface is cleaner, the reply handling is faster, and the knowledge base feature adds real value. But “better” doesn’t mean “ready.” The personalisation quality and reply handling still have meaningful gaps.
4. How does 11x compare to AiSDR or Artisan at lower price points?
AiSDR charges around $900/month with flexible monthly contracts. Artisan (Ava) is positioned at a similar price range to 11x but with multichannel capabilities including LinkedIn. At the budget end, Salesforge’s Agent Frank runs $499/month. I haven’t tested all of these head-to-head yet, but the pricing spread across AI SDRs is enormous, and the more expensive option isn’t always the best performer. An AI SDR category comparison is on my list.
5. Does 11x integrate with my CRM?
11x supports HubSpot and Salesforce natively. Other CRM integrations are available through their API or Zapier, but the depth of native integration varies. If you’re on Pipedrive, Attio, or another CRM, ask specifically about your setup during the sales process.
6. What happens to my data if I cancel?
Ask this explicitly before signing. 11x should provide a data export path for contacts, email history, and campaign analytics. Get the data retention and deletion policy in writing as part of your contract negotiation.
7. Should I wait for AI SDRs to get better, or start testing now?
If you have the budget and the right use case (high volume, large TAM, simple deal), testing now gives you a learning advantage. You’ll understand what the AI does well and where it falls short for your specific market, which means you’ll be ready to scale when the technology improves. If your budget is tight or your market is niche, waiting six months costs you nothing. The tools will be better and the pricing will likely become more competitive as the category matures.
