# Marcus Delgado, VP of Customer Success at Parcl (Series B, ~140 employees) — read of Customer Health Scorer, June 4 2026

> 9 years in CS, inherited a Gainsight license that costs more than two junior CSMs, currently trying to justify the renewal to a CFO who does not care about "health scores."

## How I got here

Googled "gainsight alternative 2026 smaller team" after our QBR where finance asked me to cut $80K in tooling. Third result was a Reddit thread, someone linked this in the comments as something they were evaluating. Opened it in a new tab, forgot about it for two days, then came back during my commute home on the Caltrain. That is genuinely how I found it.

## What I clicked first

The hero is fine. "Predict customer churn before it happens" is every CS tool ever, but the subhead does something slightly more honest: "AI-powered health scoring keeps your at-risk customers engaged and revenue stable." Revenue stable is the phrase that actually lands. I have sat in enough board meetings to know that "engaged" does not make anyone's ears perk up but "revenue stable" does. That was a small smart choice.

Then I scrolled to the "87% accuracy" claim under Predictive Accuracy. That number made me stop.

## Where I paused

"ML models trained on millions of customer behaviors predict churn with 87% accuracy."

Okay. Millions of behaviors from whose customers? What industry? What ACV range? My churn patterns at a proptech SaaS company are nothing like a horizontal HR tool or a dev tool. This number is doing a lot of work without any support. 87% sounds precise in a way that generic actually is -- it is suspiciously round-ish, suspiciously confident, and attached to zero methodology.

The FAQ says "Typical accuracy is 85-92% depending on data quality and cohort size." So the hero says 87% but the FAQ quietly widens that to a 7-point range. That delta is the tell. That is a marketing number softened in the fine print.

## What I distrusted

The bottom of the page broke the whole read for me. There is a section that says: "Honest disclosure: we don't have live customers on this idea yet. We shipped the strategy package; you ship the customer conversations."

So this is not a product. This is a product idea being sold as a landing page concept by something called Wishdeal Studio. The pricing table ($499/mo, $1,299/mo) is fake pricing for a thing that does not exist yet. The "Start Free Trial" buttons presumably go nowhere real. The "48 hours to actionable scores" claim is not from actual customer experience, it is an estimate.

I respect the disclosure existing at all. I do not respect that it requires scrolling past a full pricing table and an FAQ to find it. "Most customers see actionable scores within 48 hours" is presented as a factual customer outcome when there are no customers.

That whole middle section -- the How It Works steps, the Pricing, the FAQ -- reads like a real SaaS product. Getting to the bottom and finding out it is a pitch deck dressed as a landing page felt like a bait and switch, even if the bait-and-switcher is being honest in the fine print.

## What would convince me

If this were a real product in early access, I would want one specific case study with real numbers. Not a logo. Not a testimonial quote. Something like: "We ran this on 800 accounts at a 60-person SaaS company. Their CS team flagged 23 at-risk accounts in week one. They ran outreach on 18. 14 renewed. Here is what the health score looked like the week before the churn conversation." That is a story I can take to my CFO. "87% accuracy across millions of behaviors" is not.

If the product does not exist yet and someone is looking for a builder or early design partner, say that directly in the hero, not buried after the pricing table.

## What I'd ask in an email reply

1. The pricing table shows "Basic health scoring" on Starter and "Advanced ML models" on Growth. What is actually different between them? Because if the ML is the whole product, I need to know what I am getting at $499 that justifies not going straight to $1,299.

2. "We analyze it in your data warehouse or via API. You keep full control." Does that mean you require a data warehouse? We use Mixpanel and Salesforce but do not have Snowflake or BigQuery set up. Is that a blocker?

3. Is this live? I found the disclosure at the bottom. Are you looking for beta customers, design partners, or is there a working product I could see a demo of right now?

## Verdict: dismissive

Not of the idea, which is real and the market is real. Dismissive of this specific execution, because I cannot evaluate a product that does not exist yet from a page that presents itself as a product that does exist. If someone emailed me and said "we are building this and want two or three CS leaders to help shape it in exchange for founder pricing," I would read that email. This page does not say that.

---
*Memo by skeptic persona, generated 2026-06-04. Studio breaks own self-grading loop.*