You Have 12 Analytics Tools And No Idea What's Happening

D2C brands running $5M a year on Shopify know their numbers better than B2B businesses processing $200M.

Not because they're smarter. Because their tools just work.

Triple Whale. Northbeam. Lifetimely. God, the names just get better as I get older.

Plug in.

Connect Shopify.

See your CAC, LTV, ROAS, and blended margin on one screen.

No configuration.

No consultant.

No six-week implementation project.

Just numbers.

Meanwhile, the $200M incumbent pasted a GA4 tag on a headless site two years ago and hasn't touched it since.

Bounce rate reads 80%.

Nobody questions it.

The Monday morning dashboard says revenue is up.

That's enough, apparently.

I've had seven jobs across many years in ecommerce and B2B operations.

When I started, at each, every single one had broken analytics.

Not "could be improved" broken.

Broken broken.

Decisions made on fiction broken.

And in most cases, nobody in the room knew.

This isn't a tools problem.

It's an operations problem.

Why It Just Works, At First

Triple Whale processes 9.2 billion events daily across more than 50,000 brands. That's not a flex. That's the point. The Shopify ecosystem forced a model where analytics is an operational layer, not an IT project.

Here's what a $3M D2C brand on Shopify sees when they log into Triple Whale:

• CAC by channel, updated today
• Blended ROAS across Meta, Google, TikTok
• LTV by cohort, by acquisition source
• Margin per order after ad spend, shipping, COGS
• Creative performance down to individual ad variants

One screen.

One login.

Real numbers.

The median CPA for Meta ads across 30,000+ Triple Whale brands in 2025 was $38 and some change.

These operators know that number because they see it every morning. They don't wait for a monthly report.

They don't ask an analyst to pull a CSV.

Northbeam does the same thing for attribution modelling. Lifetimely does it for cohort LTV. The whole D2C analytics ecosystem was built by operators for operators.

Opinionated tools with opinionated defaults.

You don't configure them.

You don’t decide the important numbers.

You use them, they tell you.

But the key word. Use.

The reason D2C analytics works isn't because the tools are better engineered than GA4 or Tableau.

They probably aren't.

It's because the tools were built to be used by the person making the decision, not by a team three layers removed from the action.

A D2C founder opens Triple Whale, sees that Meta CPA spiked 22% overnight, and adjusts spend before lunch. An enterprise ecommerce manager opens GA4, sees a bounce rate that's been wrong for 18 months, and puts "analytics review" on next quarter's roadmap, telling on myself here.

Same data problem.

Completely different operational culture.

The Default Install Delusion

Here's what happens in practice at most NZ businesses above $10M in digital revenue that is not digital native.

A website gets built.

Could be headless, could be a platform, doesn't matter.

This is a glorious website, ‘Best Of Breed’ website, 6 figs website.

The agency building it includes a GA4 tag in the scope, it was part of the RFP after all.

Tag goes in during the build phase.

Maybe they set up Google Tag Manager.

Maybe they don't.

Either way, the tag fires on launch day.

Then the agency hands over the credentials and moves on to the next pitch.

That GA4 tag is now the official analytics for a business processing tens or hundreds of millions in annual revenue.

It's collecting page views.

Maybe some default scroll events.

The enhanced measurement toggles are wherever Google left them.

Dave took a look at the reports tab.

Nobody configured custom events.

Nobody mapped the actual conversion funnel.

Nobody set up cross domain tracking for the payment gateway that redirects to a third party processor.

Nobody excluded internal traffic.

Nobody checked if the consent mode implementation actually works or if it's silently dropping 40% of sessions.

But somebody, or multiple somebodys implemented what they read was best practice ignoring the 60 other tags in place.

81% of GA4 setups that get formally audited contain implementation errors that compromise data accuracy.

62.5% have misconfigured Google Tag Manager events.

These aren't edge cases.

This is the default.

And that default is what your board report is built on.

The agency that set it up? Prefacing this that most digital agencies do genuinely good work in their area of expertise. But tread lightly,because it's normally just a line item they include to complete the scope.

The people configuring your data layer are often the most junior members of the team, working from a setup guide, moving fast because the project is already over budget and two weeks behind.

The senior strategist who scoped the project and sold the analytics integration as part of the package? They're already on the next pitch before your site goes live. The handover document gets saved to a shared drive nobody opens again.

This is not malice.

This is incentive structure.

The agency is incentivised to deliver the build.

Nobody is incentivised to maintain the data layer.

And "maintain" is the operative word, because analytics isn't a one time installation. It's an ongoing operation. Tags break when the site gets updated. Events stop firing when the checkout flow changes. Cross-domain tracking fails when someone adds a new subdomain. Consent banners get updated without checking if they still pass the right signals to GA4.

The result: a beautifully designed PowerBI dashboard sitting on top of data that hasn't been accurate since launch, but thats what the board likes.

The Graveyard

The analytics stack at a mid market NZ business is genuinely impressive on paper. I've walked into companies and seen all of these tools licensed, subscribed, and technically installed:

GA4. Collecting data nobody queries beyond the default reports. Most teams can't build a custom Exploration.

GA4's most powerful features, funnel analysis, path exploration, cohort breakdowns, sit there unused. One audit found a client had been underinvesting in their highest performing channel for 14 months because the GA4 event tracking was missing 47% of lead form submissions.

Some Form of heatmap (hotjar/clarity/crazyegg). Installed. The heatmaps are live. The session recordings are running. Nobody watches them.

Ten minutes of watching real users navigate your site reveals more friction points than a month of staring at conversion rate percentages. But that requires someone whose job it is to watch. In most organisations, that person doesn't exist.

The A/B testing tool with no login. Licensed at enterprise pricing. Zero tests running.

A/B testing requires traffic volume, a hypothesis backlog, and someone to manage the testing programme.

Most companies buy the license, run one or two tests during the initial enthusiasm phase, then let it gather dust. Optimizely doesn't publish fixed pricing, but it's enterprise-level money. You're paying for a Formula 1 car and using it to drive to the shops, but they rely on you leaving it alone.

PowerBI / Tableau. Dashboards built once by a contractor or an internal BA during a reporting project. The data connections still work. The visualisations still refresh. But nobody's updated the underlying queries since the original build. The metrics on screen may or may not reflect the business as it exists today. Nobody's checking.

Any infra tool. Monitoring application performance. Tracking response times, error rates, throughput. Sending alerts that go to an inbox nobody monitors.

When the site slows down, you find out from customers complaining on a Monday morning, not from the tool that was purpose-built to tell you on Saturday night. Tracking errors nobody triages. Logging events nobody reviews. The dashboards exist. The operational discipline to use them doesn't.

Email Email flows set up during implementation. Welcome series, abandoned cart, post-purchase. Running on autopilot for 18 months. Nobody's reviewed open rates, tested subject lines, or updated the flows since setup. The tool is doing its job. The operator isn't doing theirs.

Add it up. A conservative estimate for a mid market NZ business: $10,000-80,000/year across analytics tool licenses alone.

Before you count the implementation costs, the consultant hours, and the staff time spent not using them.

That's the graveyard.

Not abandoned tools.

Technically active, operationally dead tools.

Collecting data into dashboards that refresh on schedule and inform nothing.

The BA Blind Spot

In most NZ businesses, the closest thing to an analytics function is a Business Analyst or a reporting team. These people are good at their jobs, every single one Ive met, Im in awe at how good they are at what they do.

They build dashboards.

They pull reports.

They ensure numbers are spot on.

BAs in traditional organizations report to finance, sales or ops. They're measured on their ability to tell you what happened. Revenue by channel. Units by SKU. Conversion rate as a single percentage on a Monday morning slide. Sales by region. Margin by category. Year on year comparisons.

They tell you what sold.

They almost never tell you how people buy.

‘Thats Marketings Job’.

Well, I can tell you nobody in marketing is looking at the funnel drop-offs. Nobody's watching where users hesitate, rage-click, get confused, or abandon.

Nobody's asking why 60% of traffic leaves from page two of a four-step checkout. Nobody's segmenting behaviour by device, by traffic source, by time of day, by whether the user is new or returning.

The BA knows that you converted at 2.1% last month.

The BA does not know that mobile users from paid social have an 87% cart abandonment rate because your checkout doesn't render properly on Safari, or that 34% of your desktop users click the search bar within 3 seconds of landing because your navigation is confusing, or that your highest value customers consistently enter through a category page that isn't even in your top nav.

This is the difference between a sales dashboard and an operational analytics function.

One tells you the score.

The other tells you how the game is actually being played.

D2C operators don't have this gap because their tools collapse the distinction.

Triple Whale shows acquisition and behaviour and conversion and retention in one view.

The operator sees the whole funnel.

The traditional BA sees the end of it.

You know your customers bought.

You have no idea how they buy.

But we will pave the ‘Road to $xxx’ or ensure the next target slogan is a banger, and keep it tracked on powerBI.

When D2C Goes Big

What the D2C natives hate to admit is that the model breaks at scale too.

Triple Whale is built for Shopify. It's Shopify first, Shopify native. Works beautifully when you're running a single store doing $1-20M. But the moment a D2C brand starts scaling into wholesale, retail distribution, international storefronts ala amazon, or multiple platforms, the same fragmentation problem hits.

Triple Whale doesn't natively track non Shopify revenue. Etsy, WooCommerce, Amazon, brick and mortar POS, none of it shows up in the same dashboard without a tonne of work, and by then you know its square peg into a round hole, and you look elsewhere.

Stripe transactions outside the Shopify gateway disappear. The beautiful single pane of glass becomes a window into 70% of the business, with the other 30% invisible.

Suddenly that D2C operator who had perfect visibility at $5M is flying partially blind at $50M. And the tools that got them here, the opinionated, simplified, out of the box tools, don't have the flexibility to handle the complexity.

This is the same trap, just approaching from the opposite direction. The enterprise starts with powerful tools and no operational culture. The D2C brand starts with operational culture and simplified tools. Both hit a wall when scale meets complexity.

The difference is where the gap shows up. The enterprise doesn't know what's happening on their website. The scaled D2C brand doesn't know what's happening outside of it.

Europe then entered the chat to give an additional layer of pain.

GDPR consent requirements mean GA4 in the EU is fighting with one hand tied behind its back. Consent mode gaps can silently drop tracked sessions depending on your opt in rates.

The modelling that GA4 uses to fill those gaps is better than nothing, but it's still modelling.

The US has a patchwork of state level regulations that makes consistent analytics across markets a nightmare. And everywhere, cookie deprecation continues to erode the reliability of traditional tracking.

Thanks, Apple.

The global picture is the same story at different scales. Small D2C operators have solved it with opinionated tools. Mid market businesses everywhere have an expensive tool stack and no operational layer. Enterprise businesses have entire analytics teams but still struggle with data accuracy because the underlying implementation is broken.

The problem is universal.

The question is always the same, is anyone actually using what you've already bought?

The Operator Gap

The unlock isn't buying another tool. It never was.

It's watching 50 Hotjar session recordings a week and acting on what you see. It's running 4 Optimizely tests a month and maintaining a hypothesis backlog that grows from actual user behaviour, not from someone's gut feeling in a meeting. It's building a GA4 event model that matches your real conversion funnel, not the default page_view and scroll events that shipped with the install. It's checking your New Relic alerts before your customers tell you the site is slow.

Every analytics tool has margin of error.

Every data set is imperfect.

That's fine.

The problem isn't imperfection.

It's abandonment.

D2C operators accept the noise and act anyway. They know their numbers are directionally right, and directionally right is enough to make a decision by lunchtime. Enterprise businesses wait for perfect data and get nothing. They commission analytics audits. They scope "data strategy" projects. They hire consultants to tell them what their Hotjar could show them for free if anyone would just open it.

The gap isn't technical. It's cultural. It's the difference between analytics as an operational discipline and analytics as an IT project.

So What.

Audit what you have before you buy anything new.

I guarantee you're paying for tools you've never properly used.

Start there.

Assign one human to own the data layer. Not the dashboards. The data layer. The tags, the events, the consent configuration, the cross domain tracking.

One person whose job includes making sure the data coming in is accurate.

I have nominated myself on more than one occasion.

Without this, everything built on top is fiction.

Run one tool to depth before you add a 13th.

If you have Hotjar, watch the recordings.

Every week.

If not, start for free with Clarity.

Build a process around it.

If you have A/b, start small, run tests. Maintain a backlog. Report on results.

If you have GA4, learn Explorations. Build custom funnels. Segment your traffic. Go deep before you go wide.

Stop accepting agency handovers as finished work.

The build is the beginning, not the end. Budget for ongoing analytics maintenance the way you budget for hosting.

It's infrastructure, not a project.

If your BAs can tell you what sold last Tuesday but can't tell you where customers drop off, you've built half the picture and you're making decisions on it like it's the whole thing.

The playbook isn't about Triple Whale.

It's about treating analytics like operations. Someone's job. Daily cadence. Decisions made on what the tools say, not what last month's report assumed.

The most expensive analytics tool is the one nobody opens.

Shout out to all the legends I've ripped info from for this piece:

Triple Whale | SR Analytics | Merkle | Hotjar | Northbeam