Bad Email Campaigns Don't Just Underperform – They Kill Your Audience

01 /The Concept

What if you could know how an email campaign will perform before sending a single message?

Instead of sending a real blast and hoping for the best, you run a simulation using the platform's AI engine, review the projected outcomes, and refine your copy until you have the most effective version ready to go.

The workflow is simple: upload a file with recipient addresses and email content, hit the button, and get results in under a minute. While it isn't explicitly stated, it's reasonable to assume Mocke can also connect to your existing email platform – letting it learn from actual historical send data and the behavior patterns of your specific recipients.

Simulations output projected open rates, reply rates, and unsubscribe rates for the proposed campaign.

But Mocke goes beyond what standard email analytics platforms surface. It also shows the share of recipients likely to mark the message as spam, the share likely to forward it to a colleague, the share who will set it aside to read more carefully later, and the share who will ignore it entirely.

On top of that, Mocke doesn't just show percentages – it shows the actual text of simulated replies, so you can judge how clearly your call to action landed.

One observation from the platform's own demo is worth noting: the share of recipients who simply ignored the email was roughly 20 times higher than those who marked it as spam. The real enemy of any marketing campaign isn't backlash – it's being invisible.

Critically, Mocke pairs the numbers with a written report identifying the five key factors that drove the projected results. For an outreach campaign targeting a developer tool, example factors included: "CTOs feel their engineers are fully occupied and don't want to evaluate new tools being pitched cold," and "The people receiving this email delegate tooling decisions to someone else on the team – they expect that person to already know the best options."

Mocke's creators used their own tool to test their own outreach – and claim the gap between simulated and real results averaged just 1–2%.

Pricing starts at $79 per month, scaling with the number and volume of simulations. Mocke went through Y Combinator acceleration in spring 2022 with a different product, and announced the current platform launch on Product Hunt recently.

02 /Why It Matters

Email is probably the most widely used marketing channel – but its very effectiveness depends on restraint. Send too often and you erode your list. Which means every send has to count, and the window for course correction is measured in weeks.

The fundamental problem marketers face is that when guessing how recipients will react, they tend to model their own reactions rather than their audience's. Mocke addresses this with a more objective benchmark.

Simulating user behavior more broadly is becoming a notable trend in its own right, with a cluster of startups building out the space.

Artificial Societies ([related review](/review/tema-uzhe-letit-no-vot-tak-mozhno-vzletet-povyshe)), a YC spring graduate, built a tool for creating digital twins of real audiences – so teams can test new product features, social posts, video concepts, website designs, and more against a simulated version of their actual user base. Interestingly, they got into YC partly by using their own platform: an early version simulated venture investor behavior, which they used to pressure-test their pitch before applying. The platform reportedly delivers 30% more accurate assessments than asking Claude, Gemini, or ChatGPT directly – because it simulates a specific audience rather than a generic one.

Synthetiq ([covered here](/review/hochesh-imet-kuchu-podpischikov)) launched in April to help creators predict how subscribers will respond to posts before they go live. Lakmoos ([covered previously](/review/mgnovenno-vmesto-polugoda)), a Polish startup, raised €510K to help companies quickly survey a virtual panel of potential users – with full-featured plans starting at $50,000 per year. And Velozity ([related review](/review/kak-bystro-ponjat-chto-nuzhno-polzovateljam)), launched in May, lets product teams run interviews with virtual users, including the ability to build diverse personas across different demographics and life experiences, and even run simulated focus groups.

03 /Opportunities

The overarching trend here is using AI to forecast the effectiveness of marketing and product decisions before committing to them.

The key differentiator is specificity: platforms that model a genuine digital replica of a particular audience will consistently outperform generic AI prompts. The winners in this space will be those that build the most accurate and nuanced audience simulations.

The most defensible play in this space: build the most accurate replica of a specific audience first – your own. The competitive edge comes from specificity that generic AI can't replicate, which means the internal tool often becomes the product.