The Dashboard Obsession Is Killing Your Business

Somewhere along the way, the business world decided that dashboards equal data maturity. Want to look sophisticated? Build a dashboard. Want to impress the board? Show them a dashboard. Want to feel like you're in control? Open a dashboard.

Here's the problem: a beautiful dashboard built on bad data is worse than no dashboard at all.

No dashboard means you know you're flying blind. A bad dashboard means you think you can see — but you're looking at a mirage.

I've audited over 200 HubSpot portals. In at least 80% of them, the reporting was unreliable. Not because HubSpot's reporting tools are bad (they're actually quite good). Because the data underneath was broken.

And nobody was checking.

The 5 Data Quality Problems Hiding in Every HubSpot Portal

After 200+ implementations, these are the five problems I find every single time. Not sometimes. Every time.

Problem 1: Inconsistent Deal Amounts

The most common — and most expensive — data quality issue.

Sales rep A enters deal amounts including GST. Sales rep B enters them excluding GST. Sales rep C enters a "best guess" and never updates it. The finance team enters the actual invoice amount — which is different from all three.

Your pipeline report adds them all together and shows you a number that means absolutely nothing.

The fix: One standardised rule. Is it ex-GST or inc-GST? Document it. Make it a required field with a helper text that says exactly what to enter. Build a validation workflow that flags deals where the amount changes by more than 20% (likely an error).

Problem 2: Deal Stage Definitions That Nobody Follows

"Qualified" means something different to every person on your team.

For your best rep, "Qualified" means they've had a discovery call, confirmed budget, and identified the decision-maker. For your newest rep, "Qualified" means the prospect replied to an email.

Your pipeline report treats them identically. So your conversion rate from "Qualified" to "Closed Won" is meaningless — because you're comparing apples to office furniture.

The fix: Write a one-sentence definition for every deal stage. Print it out. Stick it on the wall. Build it into HubSpot as stage requirements. A deal can't move to "Qualified" unless Budget, Authority, and Timeline fields are filled in.

Problem 3: No Source Attribution (Or Wrong Source Attribution)

"Where do our best clients come from?"

If you can't answer that question with confidence, your marketing spend is a guess.

Most HubSpot portals I audit have 40-60% of contacts with no original source — or worse, the wrong source. Contacts created manually default to "Offline Sources." Imported lists default to "Other." Someone who found you on Google, called your office, and got manually added? HubSpot thinks they're "Offline."

The fix: Create a required "Lead Source" property that's set at the point of entry. Build workflows that automatically set source based on form submissions, ad clicks, and referral URLs. For manual entries, make source a required field — no exceptions.

Problem 4: Duplicate Contacts Everywhere

I opened a client's portal last month. 28,000 contacts. After deduplication: 19,000. That's 9,000 duplicates — 32% of their entire database.

Every duplicate skews your reporting. Email open rates are wrong (same person counted twice). Deal attribution is split. Lifecycle stage reporting is unreliable. Your "1,000 new leads this quarter" might actually be 700 new leads and 300 duplicates.

The fix: Run HubSpot's built-in deduplication tool monthly. Set up workflows that check for duplicates at the point of creation (match on email address). For companies, merge based on domain. Schedule a quarterly data cleanup — put it in the calendar, make it non-negotiable.

Problem 5: Properties That Exist But Nobody Fills In

The average HubSpot portal I audit has 200-400 custom properties. When I ask "which of these are actually being used?" the answer is usually: "I'm not sure."

Empty properties aren't just clutter — they're broken reporting dimensions. You built a "Customer Segment" property with great intentions. 15% of contacts have it filled in. You build a report segmented by Customer Segment. The report shows 85% of your contacts as "Unknown." Useless.

The fix: Audit every custom property. If less than 50% of relevant records have it populated, either delete the property or build a workflow to populate it. No middle ground. Simplicity scales. Complexity dies.

The Data Before Dashboards Framework: 5 Steps

Here's the exact process we follow with every client before we build a single report.

Step 1: Audit Your Data Sources (Day 1)

List every way data enters your HubSpot: forms, imports, manual entry, integrations, API connections. For each source, answer: Is there validation? Is there standardisation? Is there deduplication?

Most businesses have 5-8 data entry points. At least half will have no quality controls.

Step 2: Define Your Core Metrics (Day 2)

Before you build dashboards, answer: What are the 5 numbers that actually run this business?

Not 50. Not 25. Five.

For most B2B businesses, it's:

• Pipeline value (total open deals)
• Close rate (deals won / deals created)
• Average deal size
• Time to close
• Revenue by source

Step 3: Clean the Foundation (Days 3-10)

This is where the real work happens. It's not glamorous. It's data cleaning.

• Deduplicate contacts and companies
• Standardise deal amounts (ex-GST or inc-GST, pick one)
• Define and enforce deal stage criteria
• Fill in or delete empty custom properties
• Fix source attribution for at least the last 12 months

I won't sugarcoat it: this takes time. For most portals, 5-8 business days of focused work. But it's a one-time investment that makes everything after it trustworthy.

Step 4: Build Quality Enforcement (Days 11-14)

Now build the guardrails so the data stays clean:

• Required fields at each deal stage
• Validation workflows that flag anomalies
• Automatic deduplication on contact creation
• Source attribution workflows for every entry point
• A monthly data quality check (30 minutes, put it in the calendar)

This is the step most businesses skip. They clean the data once and then wonder why it's messy again 3 months later. Data quality isn't a project. It's a discipline.

Step 5: NOW Build Your Dashboards (Days 15+)

Only now — after the data is clean, standardised, and protected — do you build reports.

And here's the beautiful thing: when your data is clean, you don't need 47 custom dashboards. You need 3-5 simple reports that show your core metrics. They'll be accurate. You'll trust them. And you'll actually use them.

Data before dashboards. Always.

The Real Cost of Bad Data

Let me make this concrete.

A $2M/year services business with dirty CRM data is likely experiencing:

Lost revenue from poor lead attribution — ~$80K/year If you don't know where your best clients come from, you're spending marketing dollars in the wrong places. That client who discovered they got 42% of business from referrals and only 12% from paid ads? They were spending $4K/month on ads that barely moved the needle. That's $48K/year misallocated — and the opportunity cost of not investing in referral programs is even higher.

Lost revenue from slow follow-up — ~$60K/year When your pipeline data is unreliable, leads fall through cracks. Deals sit in wrong stages. Follow-up tasks don't fire. In our experience, fixing pipeline data quality alone improves close rates by 10-15%.

Wasted time on manual reporting — ~$30K/year When your dashboards are wrong, someone builds spreadsheets to "double-check." I've seen ops managers spend 8 hours a week manually reconciling CRM data with accounting data. That's a $30K+/year salary cost spent on work that shouldn't exist.

Decision-making cost — incalculable This is the big one. How do you quantify six months of decisions based on a $150K error? You can't. But you know it's expensive.

Your 10-Minute Data Quality Check

You don't need to hire us to start. Here's a quick self-assessment you can do right now.

Open your HubSpot. Answer these 5 questions:

1. Pull up your last 10 closed-won deals. Do the amounts match your invoices? If more than 2 are off by more than 10%, you have a deal amount standardisation problem.

2. How many custom properties do you have? How many have >80% fill rate? Go to Settings → Properties → Export. Count them. If more than half have less than 80% fill rate on relevant records, you have a property bloat problem.

3. Run the deduplication tool. How many duplicates does it find? If it's more than 5% of your database, you have a duplicate problem.

4. Pull up 20 random contacts. Do they all have an original source? If more than 3 are missing or show "Offline Sources" when they shouldn't, you have an attribution problem.

5. Ask your two best sales reps to define "Qualified." Do they give the same answer? If not, your pipeline stages aren't standardised. Score:

• 0-1 issues: Your data is in decent shape. Build those dashboards.
• 2-3 issues: You need a cleanup before any reporting project.
• 4-5 issues: Stop everything. Fix the data first. Those dashboards you're relying on? They're lying to you.
  
  

The Bottom Line

I'm not anti-dashboard. I love a good report. There's nothing more satisfying than opening a dashboard and seeing accurate, real-time business metrics that you can trust.

The keyword is trust.

You can't trust a dashboard built on dirty data. And most businesses have dirtier data than they think.

So before you spend another dollar on reporting tools, custom dashboards, or data visualisation — ask yourself: is the data underneath worth visualising?

If the answer is "I'm not sure" — you know where to start.

Data before dashboards. Always.