Enginy claims AI sales tools save 20 hours per SDR every week and cut cost of acquisition by 70%. Those numbers are aggressive. Most vendors quote them. Few can prove them against your pipeline. The scale of potential savings justifies a serious evaluation, but you need a framework to separate real gains from vendor theatre.
What the 20-hour claim actually means
Twenty hours per SDR per week is half a standard work week. That is not optimisation. That is a different job. For a solo founder running outbound, that is the difference between prospecting and doing the work you actually sell.
Break the claim down and it becomes more plausible. A typical SDR spends roughly 30% of their week on research, 20% on writing, and 15% on data entry. AI tools that automate inbox scoring, draft personalisation, and meeting prep attack those blocks directly. The math works if the tool is good and the workflow is built around it.
But here is the catch. The 20-hour figure assumes your SDR was already doing those tasks manually at full speed. Most founders we talk to are not. They are skipping research, sending generic templates, and losing deals in follow-up. The tool does not save time you were never spending. It improves quality first, speed second.
Four times the meetings is a red flag, not a promise
Enginy also claims 4x meetings booked. That number should make you suspicious. If a tool genuinely quadrupled meetings for every user, every sales team on earth would have adopted it by now. They have not. The claim is likely based on a specific segment, a specific ICP, or a specific baseline that was already underperforming.
What we have seen in practice is more modest. A customer running outbound to UK accountancy firms went from 11 meetings per month to 23 after switching from manual research to an AI-assisted workflow. That is 2.1x, not 4x. The difference was not the AI. It was the consistency. They sent 40 emails a week instead of 18 because the research bottleneck disappeared.
Treat 4x as an upper bound, not an expectation. If you get 2x, you are doing well. If you get 1.5x, the tool is still paying for itself if it saves you the research time.
How to validate the 70% lower CAC claim
Cost of acquisition is the only number that matters at the end of the quarter. The 70% reduction claim is plausible but only under specific conditions. It works when your CAC is dominated by SDR labour, not by ad spend or content marketing.
Here is the validation framework we use with founders:
- Calculate your current CAC per meeting booked, not per customer. Include salary, tooling, and data costs.
- Run a two-week pilot with one AI tool on one ICP segment. Keep everything else identical.
- Compare meetings booked per hour of SDR time, not total meetings. That is the efficiency metric.
- Extrapolate only if the pilot shows at least 1.5x improvement. Anything less is noise.
We ran this exact test with a B2B SaaS founder selling at $30K ACV. His CAC per meeting dropped from £180 to £95. That is 47%, not 70%. The gap came from his own time spent reviewing AI-generated emails before sending. He refused to send anything he had not edited. That review time is real and it is not going away.
The human-in-the-loop cost nobody budgets for
Every AI sales platform promises automation. None of them should remove the human from the send button. The review loop is where the quality lives, and it is also where the time savings shrink.
We built MiraReach with that constraint baked in. The platform drafts, scores, and prepares. It never sends. That means the 20-hour claim is not achievable for us either, because we are not trying to replace the SDR. We are trying to make the SDR's judgment faster and better informed.
Budget for 15 to 20 minutes per day of review time per SDR. That is the cost of not embarrassing your domain. If a vendor tells you zero review is needed, they are selling spam, not sales.
What to ask before you buy any AI sales tool
Before you sign up for anything, ask these four questions. The answers will tell you more than any case study.
First, what is the baseline for your efficiency claims? If they cannot tell you the exact workflow and ICP of the customers in the case study, the numbers are marketing. Second, can you run a two-week pilot on my data, not your demo data? If the answer is no, walk away. Third, what happens to my data if I cancel? Some tools hold your prospect lists hostage. Fourth, does the tool integrate with my existing stack or do I need to rebuild my workflow around it? The latter is a hidden cost that eats your savings.
We have seen founders burn two weeks migrating from one AI tool to another because the first one could not export cleanly. That is not efficiency. That is a tax.
What we would do next
Run the pilot. Pick one ICP, one tool, two weeks. Measure meetings per hour, not total meetings. If you see 1.5x or better, scale it. If you do not, the problem is probably your offer, not the tool. AI cannot fix a weak value proposition.
If you want to test this framework with a tool that keeps you in the loop, see how MiraReach handles this. We built it for founders who want the efficiency without losing control of the message.
— Mira