How We Built a Platform That Did Everything (and Nothing)

In 2024, we started building Polynym as a multi-brand management and growth platform.

The pitch: one dashboard to track sales, leads, and followers across all your brands. Monitor revenue, attribute lead sources, get AI-powered insights, forecast performance, and see everything in real-time.

We launched with a free tier (3 brands, 100 months of data) and a Pro plan (£49/month, unlimited brands, AI insights). By mid-2025, we claimed 500+ brands tracked and 10M+ data points processed.

By November, we shut it down.

The Original Vision

The idea came from watching agencies and multi-brand businesses drown in dashboards.

An agency contact showed me her Monday routine: Check Shopify sales for Brand A. Check Instagram followers for Brand B. Check Facebook Ads for Brand C. Export three spreadsheets. Make charts. Email clients.

"I spend half my week just pulling numbers," she said.

So we built Polynym. One unified dashboard that would show:

• Multi-brand performance tracking
• Sales and revenue analytics
• Social media follower counts across platforms
• Lead source attribution
• AI-powered insights
• Performance forecasting
• Custom reporting

Pull in all your brands, see everything in one place, get AI recommendations on what to focus on next. Free for 3 brands, £49/month for unlimited brands and AI features.

It seemed obvious.

The Feature Creep Trap

Month 3: "Can it track TikTok follower growth too?"

Month 4: "What about email marketing metrics?"

Month 5: "Can the AI predict which products will sell best next quarter?"

Month 6: "We need inventory alerts."

Month 8: "What about customer support ticket tracking?"

Every feature request made sense. Every one solved a real problem for someone. So we kept adding them.

By November 2025, Polynym could:

• Track 9 social platforms
• Monitor 14 data sources
• Generate 6 types of reports
• Send 12 kinds of alerts
• Create 4 types of forecasts
• Integrate with 23 services

People who tried it asked, "What does it actually do?"

Good question.

The Honest Numbers

November 2025. Stats we showed on the website:

• 500+ brands tracked
• 10M+ data points processed
• 99.9% uptime

Stats we didn't show:

• 23 beta users
• 12% retention rate
• 4 minute average session time
• 0 conversions from free to £49/month Pro

The feedback was consistent:

• "It's too much."
• "I don't need half these features."
• "I still don't know what problem it solves."

One user: "I came for Instagram follower tracking. Why is there a revenue forecasting tab with machine learning settings?"

We had built a platform that could track everything for everyone. Which meant it did nothing particularly well for anyone.

The dashboard had 8 navigation items. The settings page had 112 toggles. The AI gave insights about metrics users didn't care about. The onboarding took 15 minutes and people still didn't understand what to look at first.

£49/month for a dashboard nobody opened more than once.

The Pivot Decision

The obvious move was to cut features. Keep the core, delete the rest, relaunch as "Polynym 2.0 - Simplified."

But which features were the core? Social media tracking? Sales analytics? Lead attribution? Forecasting?

Different users gave different answers. Which meant there was no core. Every feature was essential to someone and useless to everyone else.

We had two options:

1. Pick one thing, delete the rest, compete with established tools in that category
2. Shut down the platform, build focused products instead

We chose option 2.

If Instagram analytics was valuable for some users and inventory management was valuable for different users, why force them into the same product? Why make social media managers navigate through revenue forecasting tabs they'll never touch?

Focus beats features. Four focused products beat one unfocused platform.

What Polynym Is Now

Polynym is a UK product studio. We build small, focused SaaS tools:

PropertyScribe: AI property descriptions for UK estate agents. Upload photos of a property, get a complete listing description in 30 seconds. Does one thing.

DailyExtract: Daily email summary of specific Slack channels. Tell it which channels to watch, get one email every morning with what matters. Does one thing.

BugFeed: Bug reporting widget for websites. Users click the widget, it auto-captures a screenshot and device info, creates a ticket. No more "what browser were you using?" Does one thing.

StackPod: Run databases locally without Docker. One click to start PostgreSQL, Redis, or MongoDB. No YAML files. No configuration. Does one thing.

Four products. Four problems. Four clear solutions.

Nobody asks "What does PropertyScribe do?" because it only does one thing.

We also offer productised services now (Workflow Automation, AI Solutions, Custom Platforms, SaaS Development) but each service has a clear scope and outcome.

What We Learned

1. More features don't make a better product

We thought adding AI-powered insights and forecasting would make Polynym worth £49/month. It made it more confusing.

The best products do one thing so well that you don't need anything else for that use case.

2. "Agencies and multi-brand businesses" is not a target market

When we said Polynym was for "agencies and multi-brand businesses," we were really saying "we don't know who this is for."

Generic products get generic results. 12% retention results.

3. Admitting failure is cheaper than pretending it works

We could have spent another year tweaking Polynym. Better onboarding. Simpler UI. More AI features. Ads targeting agencies.

But we knew it wasn't working. Admitting that in November instead of May saved us six months.

4. Pivoting doesn't mean starting over

We didn't throw away the code. PropertyScribe uses the AI content engine from Polynym. BugFeed uses the notification system. DailyExtract uses the email infrastructure and data processing.

The parts were fine. Forcing them into one platform was the problem.

5. Focus is a feature

When PropertyScribe launched, we got emails asking if we'd add commercial property support, or rental descriptions, or social media captions for estate agents.

We said no to all of them. Not because they're bad ideas, but because they're different products.

Saying no is how products stay focused.

The New Rule

Before building anything now, we ask: "Can we describe what this does in one sentence without using 'and'?"

PropertyScribe: "Generates property descriptions for estate agents." DailyExtract: "Sends daily email summaries of Slack channels." BugFeed: "Bug reporting widget with auto-capture." StackPod: "Runs databases locally without Docker."

If we need "and" to describe it, we're building a platform again.

Platforms are where products go to die.

---

Update (January 2026): All four products are now live. PropertyScribe, DailyExtract, BugFeed, and StackPod. Each does one thing. Each has paying customers.

Four focused products beat one unfocused platform with 500+ brands tracked and 12% retention.