KompenaBETA
OverviewRecommendationsAnalysesInsightsStrategyAssistant
Budget plannerNotes
Data sources
Google Analytics 4✓
Google Search Console✓
Google Ads✓
Meta Ads✓
Klaviyo!
Shopify✓
Connect more →
NCNordic Coffee ApSBETA
OverviewRecommendationsRoadmapInsightsStrategyAssistantBudget plannerNotesData sources

Analyses

Last updated 10 Jul 2026, 09:42
Upgrade

Every question we know how to ask of your numbers, gathered in one place. An analysis is a spec rather than code: it names the metrics it reads, the thresholds it holds them against and the recommendations it may raise. Nothing runs until you say so, and every answer traces back to a number.

53 in the library

SEO

7 analyses
Are people skipping you in the search results?You may rank and still be ignored. This compares your organic click-through rate against the 2% floor where snippets usually need rewriting.
SEO · Built in, by us

No money ceiling on this one, because the recommendations it can raise have no deterministic formula on the data we hold. A made-up figure would be worse than none.

Are you only found by people who already know you?Splits your organic clicks into people who searched your name and people who searched for what you sell. Only the second group can become new customers.
SEO · Built in, by us

No money ceiling on this one, because the recommendations it can raise have no deterministic formula on the data we hold. A made-up figure would be worse than none.

Is your organic traffic quietly shrinking?Compares organic clicks over the last 30 days against the 30 before. A slow decline is the most expensive thing in search, because nothing visibly breaks.
SEO · Built in, by us

No money ceiling on this one, because the recommendations it can raise have no deterministic formula on the data we hold. A made-up figure would be worse than none.

Is anyone searching for your name at all?Branded search is the cleanest read on whether the market remembers you. It moves in search data long before it moves in the sales figures.
SEO · Built in, by us

No money ceiling on this one, because the recommendations it can raise have no deterministic formula on the data we hold. A made-up figure would be worse than none.

You are being shown. Are you being chosen?Impressions mean Google thinks you are relevant. Clicks mean the searcher agreed. This measures the distance between the two, and whether there is enough volume for that distance to be worth money.
SEO · Built in, by us

No money ceiling on this one, because the recommendations it can raise have no deterministic formula on the data we hold. A made-up figure would be worse than none.

Can customers in your own area find you?For a service business most demand is local and decided in the map results, where review volume and a complete profile do most of the ranking.
SEO · Built in, by us

No money ceiling on this one, because the recommendations it can raise have no deterministic formula on the data we hold. A made-up figure would be worse than none.

Can Google load and read your pages properly?We cannot see your index coverage from outside Search Console, so this checks the part we can measure: whether the pages load fast enough and are structured cleanly enough to compete.
SEO · Built in, by us
Could uncover up to ~15k kr/mo

That is a ceiling, not a promise: it sums what the recommendations behind this analysis would be worth on your own numbers if every check fired, and until it runs nobody knows whether anything is wrong at all. 1 of the recommendations it can raise have no money formula, so they are not in that number.

Paid

9 analyses
Does your ad spend earn its keep?Blended ROAS under 3 means most of the revenue is going back out the door as media cost.
Paid search · Built in, by us

No money ceiling on this one, because the recommendations it can raise have no deterministic formula on the data we hold. A made-up figure would be worse than none.

Is your next krone of ad spend still buying growth?Average ROAS flatters the last krone you spent. This applies the same diminishing-returns curve the budget optimizer uses (revenue = spend^e) to work out what the NEXT krone actually returns, per channel.
Paid search · Built in, by us

No money ceiling on this one, because the recommendations it can raise have no deterministic formula on the data we hold. A made-up figure would be worse than none.

Do you have room to spend more before returns flatten?The same spend^e curve that warns about saturation also shows headroom. This checks whether your marginal krone is still comfortably above break-even, which is the only honest case for raising a budget.
Paid search · Built in, by us

No money ceiling on this one, because the recommendations it can raise have no deterministic formula on the data we hold. A made-up figure would be worse than none.

Is your budget on the wrong side of the Google/Meta split?Two channels, one budget. This compares what each one actually returns, so the split is decided by numbers rather than by habit or by whichever platform reports the friendliest figure.
Paid search · Built in, by us

No money ceiling on this one, because the recommendations it can raise have no deterministic formula on the data we hold. A made-up figure would be worse than none.

How much of your traffic rests on a single channel?Concentration is cheap until the day it is not: one algorithm change, one account suspension, one CPM spike. This measures the largest channel's share of your sessions, which is the closest read we have on how exposed the budget is.
Paid search · Built in, by us

No money ceiling on this one, because the recommendations it can raise have no deterministic formula on the data we hold. A made-up figure would be worse than none.

Are you renting traffic you could own?Paid traffic stops the day the card is declined; organic does not. This puts your paid visibility next to your organic visibility on the same 0 to 100 scale, and checks whether the free side is shrinking.
Paid search · Built in, by us

No money ceiling on this one, because the recommendations it can raise have no deterministic formula on the data we hold. A made-up figure would be worse than none.

Is the paid click worth less than your other clicks?When creative gets tired or the audience is exhausted, the click still costs full price but converts worse. This compares paid traffic against your own site average, and checks whether the ads promise something the landing page does not deliver.
Paid social · Built in, by us
Could uncover up to ~19k kr/mo

That is a ceiling, not a promise: it sums what the recommendations behind this analysis would be worth on your own numbers if every check fired, and until it runs nobody knows whether anything is wrong at all. 1 of the recommendations it can raise have no money formula, so they are not in that number.

Does the bidding algorithm know which clicks became sales?Automated bidding spends your budget on the conversions it can see. This checks how much of your purchase data actually reaches the platforms server-side, because a blind algorithm buys the wrong clicks at full price.
Paid social · Built in, by us

No money ceiling on this one, because the recommendations it can raise have no deterministic formula on the data we hold. A made-up figure would be worse than none.

Is paid spend covering for demand that is drying up?Ads can harvest existing demand or create new demand, and the two cost very different amounts. This checks whether interest in your name is falling, which is what forces paid to work harder every month for the same result.
Paid social · Built in, by us

No money ceiling on this one, because the recommendations it can raise have no deterministic formula on the data we hold. A made-up figure would be worse than none.

Email

9 analyses
Is email pulling its weight?For most shops email should carry a fifth of revenue. This checks what share yours actually brings in.
Email & CRM · Built in, by us

No money ceiling on this one, because the recommendations it can raise have no deterministic formula on the data we hold. A made-up figure would be worse than none.

How many started checkouts never finish?Abandonment above 70% usually means the recovery flow is missing rather than that the customers changed their minds.
Email & CRM · Built in, by us

No money ceiling on this one, because the recommendations it can raise have no deterministic formula on the data we hold. A made-up figure would be worse than none.

Is email selling, or only sending?Email is the one channel that writes to people who already know you. This checks whether it produces revenue at all, against the 5% where our email score bottoms out.
Email & CRM · Built in, by us

No money ceiling on this one, because the recommendations it can raise have no deterministic formula on the data we hold. A made-up figure would be worse than none.

Is too much of the month riding on your list?A high email share looks like strength until you notice what it means: most revenue comes from people you already have. This flags when the list is carrying acquisition's job.
Email & CRM · Built in, by us

No money ceiling on this one, because the recommendations it can raise have no deterministic formula on the data we hold. A made-up figure would be worse than none.

Do enough visitors join the list?Every flow you build only reaches the addresses you captured. This measures signup rate against the 2% a visible signup offer usually clears, and against how often visitors buy.
Email & CRM · Built in, by us

No money ceiling on this one, because the recommendations it can raise have no deterministic formula on the data we hold. A made-up figure would be worse than none.

Are the standard email flows switched on?We count four flows that need no new traffic: welcome, abandoned cart, browse abandonment and win-back. This checks how many are actually live in your email platform.
Email & CRM · Built in, by us

No money ceiling on this one, because the recommendations it can raise have no deterministic formula on the data we hold. A made-up figure would be worse than none.

Is anything catching the checkouts you nearly won?An abandoned checkout is a customer who picked a product and stopped at payment. This sizes that volume against your completed orders, and checks how far the checkout itself falls short.
Email & CRM · Built in, by us

No money ceiling on this one, because the recommendations it can raise have no deterministic formula on the data we hold. A made-up figure would be worse than none.

Does anything mail your customers between orders?A second order costs nothing to acquire. This one only runs when we know your order volume, and asks whether the flows aimed at people who already bought are running at all.
Email & CRM · Built in, by us

No money ceiling on this one, because the recommendations it can raise have no deterministic formula on the data we hold. A made-up figure would be worse than none.

Can we even see a way to join your list?Before any email platform is connected, we look at your own front page for a signup form and a recognizable email platform behind it, and score it 0, 50 or 100.
Email & CRM · Built in, by us

No money ceiling on this one, because the recommendations it can raise have no deterministic formula on the data we hold. A made-up figure would be worse than none.

CRO

11 analyses
Do you lose mobile visitors that desktop keeps?Most traffic is mobile. This compares the two conversion rates directly, so a mobile problem can't hide inside an average.
Conversion · Built in, by us
Could uncover up to ~15k kr/mo

That is a ceiling, not a promise: it sums what the recommendations behind this analysis would be worth on your own numbers if every check fired, and until it runs nobody knows whether anything is wrong at all. 1 of the recommendations it can raise have no money formula, so they are not in that number.

Do your product pages convince?Measures add-to-cart rate on product pages against the 10% benchmark, so you know whether the problem is traffic or the page.
Conversion · Built in, by us

No money ceiling on this one, because the recommendations it can raise have no deterministic formula on the data we hold. A made-up figure would be worse than none.

Does your site convert like the rest of your market?Compares your site-wide conversion rate with the 2% we use for Danish webshops, so you can tell whether the problem is how much traffic you buy or what the site does with it.
Conversion · Built in, by us
Could uncover up to ~19k kr/mo

That is a ceiling, not a promise: it sums what the recommendations behind this analysis would be worth on your own numbers if every check fired, and until it runs nobody knows whether anything is wrong at all. 1 of the recommendations it can raise have no money formula, so they are not in that number.

Is mobile carrying its own weight, whatever desktop does?A mobile-versus-desktop comparison goes quiet when both are weak. This one holds mobile up against the market prior on its own, because most of your visitors are on a phone and that rate sets your total.
Conversion · Built in, by us
Could uncover up to ~15k kr/mo

That is a ceiling, not a promise: it sums what the recommendations behind this analysis would be worth on your own numbers if every check fired, and until it runs nobody knows whether anything is wrong at all. 1 of the recommendations it can raise have no money formula, so they are not in that number.

How large is the mobile gap, in relative terms?Knowing that mobile trails desktop is not enough to prioritize it. This measures how much of desktop's conversion rate mobile is missing, and says where that number stops being ordinary device behaviour.
Conversion · Built in, by us
Could uncover up to ~15k kr/mo

That is a ceiling, not a promise: it sums what the recommendations behind this analysis would be worth on your own numbers if every check fired, and until it runs nobody knows whether anything is wrong at all. 1 of the recommendations it can raise have no money formula, so they are not in that number.

Do the visitors you pay for convert worse than the free ones?Paid traffic arrives with an expectation set by the ad. If it converts below your own site average, the landing page is not answering the promise the ad made.
Conversion · Built in, by us
Could uncover up to ~34k kr/mo

That is a ceiling, not a promise: it sums what the recommendations behind this analysis would be worth on your own numbers if every check fired, and until it runs nobody knows whether anything is wrong at all.

How many finish checkout once they have started it?Counts completed purchases as a share of started checkouts against the 48 to 52% we see for Danish webshops. This is your most expensive drop-off, because these people had already chosen a product.
Conversion · Built in, by us

No money ceiling on this one, because the recommendations it can raise have no deterministic formula on the data we hold. A made-up figure would be worse than none.

Is page speed taxing every conversion you make?Core Web Vitals count how many of Google's three field thresholds (LCP, CLS, INP) your real visitors actually pass. Google's bar is all three, and a failing one is friction on every page, not just one.
Conversion · Built in, by us
Could uncover up to ~15k kr/mo

That is a ceiling, not a promise: it sums what the recommendations behind this analysis would be worth on your own numbers if every check fired, and until it runs nobody knows whether anything is wrong at all. 1 of the recommendations it can raise have no money formula, so they are not in that number.

Are the mechanical basics of the homepage in place?A crawl checks six fundamentals: https, a mobile viewport tag, a usable title, a meta description, one h1, and structured data. It cannot tell you whether the site is pleasant to use, only whether the plumbing is there.
Conversion · Built in, by us
Could uncover up to ~19k kr/mo

That is a ceiling, not a promise: it sums what the recommendations behind this analysis would be worth on your own numbers if every check fired, and until it runs nobody knows whether anything is wrong at all. 1 of the recommendations it can raise have no money formula, so they are not in that number.

Can a first-time visitor tell what you sell and why you?This is an assessed reading of your hero message on clarity and differentiation, not a measurement of behaviour, and we say so because you should be able to disagree with it.
Conversion · Built in, by us
Could uncover up to ~19k kr/mo

That is a ceiling, not a promise: it sums what the recommendations behind this analysis would be worth on your own numbers if every check fired, and until it runs nobody knows whether anything is wrong at all.

Do you keep the visitors who were not ready to buy?Most first-time visitors will not buy today. This checks what share of new visitors leave you a way to reach them again, against the 2% we set as the floor for a site that actively asks.
Conversion · Built in, by us
Could uncover up to ~19k kr/mo

That is a ceiling, not a promise: it sums what the recommendations behind this analysis would be worth on your own numbers if every check fired, and until it runs nobody knows whether anything is wrong at all. 1 of the recommendations it can raise have no money formula, so they are not in that number.

Tracking

8 analyses
Is the ad platforms' view of your sales complete?Meta and Google optimize on the conversions they can see. This checks how much of your purchase data actually reaches them.
Tracking · Built in, by us

No money ceiling on this one, because the recommendations it can raise have no deterministic formula on the data we hold. A made-up figure would be worse than none.

Does Google still see the visitors who reject cookies?Consent Mode decides what Google is allowed to receive and what it may model. This checks how far your setup actually goes: nothing, basic, or advanced.
Tracking · Built in, by us

No money ceiling on this one, because the recommendations it can raise have no deterministic formula on the data we hold. A made-up figure would be worse than none.

Do your tags survive ad blockers and Safari?Browser tags are the first thing an ad blocker stops and the first thing Safari puts a time limit on. This checks whether you have a server container standing behind them.
Tracking · Built in, by us

No money ceiling on this one, because the recommendations it can raise have no deterministic formula on the data we hold. A made-up figure would be worse than none.

Does Meta get a server-side copy of every conversion?Meta's redundant setup expects every browser event to have a twin sent from your server. This checks how much of your data actually arrives that way, and how well Meta can match it to a person.
Tracking · Built in, by us

No money ceiling on this one, because the recommendations it can raise have no deterministic formula on the data we hold. A made-up figure would be worse than none.

Is Google Ads bidding on conversions it can trust?Smart Bidding is only as good as the conversion it is handed. This checks whether your Google Ads conversions are recording at all, and whether they are verified and enhanced.
Tracking · Built in, by us

No money ceiling on this one, because the recommendations it can raise have no deterministic formula on the data we hold. A made-up figure would be worse than none.

Can your analytics check the ad platforms' numbers?If GA4 has no reliable purchase funnel of its own, there is no second number to hold Meta and Google up against. This checks the events that funnel is built on.
Tracking · Built in, by us

No money ceiling on this one, because the recommendations it can raise have no deterministic formula on the data we hold. A made-up figure would be worse than none.

Do your analytics and Search Console tell the same story?Two independent sources count the same organic visits, and only one of them depends on a tag firing. This compares them directly.
Tracking · Built in, by us

No money ceiling on this one, because the recommendations it can raise have no deterministic formula on the data we hold. A made-up figure would be worse than none.

Which channel takes credit for another channel's work?Last-click reporting hands the sale to the final touch. This looks for the two patterns where that reliably points you at the wrong channel.
Tracking · Built in, by us

No money ceiling on this one, because the recommendations it can raise have no deterministic formula on the data we hold. A made-up figure would be worse than none.

Brand

7 analyses
Is there enough proof that others trust you?Below a 4.0 average, reviews stop reassuring buyers and start costing you the sale.
Brand · Built in, by us

No money ceiling on this one, because the recommendations it can raise have no deterministic formula on the data we hold. A made-up figure would be worse than none.

Are there enough reviews for your rating to mean anything?A good average on ten reviews is not proof. This checks the volume behind your rating, because volume is what makes a rating hard to argue with.
Brand · Built in, by us

No money ceiling on this one, because the recommendations it can raise have no deterministic formula on the data we hold. A made-up figure would be worse than none.

Would your reviews win a side-by-side comparison?Rating and volume together, the way a buyer sees them with two profiles open in two tabs. A 4.6 on 12 reviews loses to a 4.3 on 900.
Brand · Built in, by us

No money ceiling on this one, because the recommendations it can raise have no deterministic formula on the data we hold. A made-up figure would be worse than none.

Are fewer people looking you up by name?Branded search is the closest thing to a demand meter you own. This checks whether it is growing, and whether it keeps up with the rest of your traffic.
Brand · Built in, by us

No money ceiling on this one, because the recommendations it can raise have no deterministic formula on the data we hold. A made-up figure would be worse than none.

Is your organic traffic demand you own or demand you rent?The share of organic clicks that comes from your own name, checked in both directions: too little means nobody knows you, too much means organic is only harvesting people who already do.
Brand · Built in, by us

No money ceiling on this one, because the recommendations it can raise have no deterministic formula on the data we hold. A made-up figure would be worse than none.

Are you visible where your competitors advertise?Meta's and Google's public ad libraries show what anyone is running, you and them. This checks how much of your brand is actually out there, and whether it says the same thing twice.
Brand · Built in, by us
Could uncover up to ~19k kr/mo

That is a ceiling, not a promise: it sums what the recommendations behind this analysis would be worth on your own numbers if every check fired, and until it runs nobody knows whether anything is wrong at all. 1 of the recommendations it can raise have no money formula, so they are not in that number.

Does your own site give a stranger a reason to believe you?Reviews live off-site, but the reasons to trust you have to be on it. This checks whether the homepage says what you sell, and whether the credibility markers a machine can read are there at all.
Brand · Built in, by us
Could uncover up to ~19k kr/mo

That is a ceiling, not a promise: it sums what the recommendations behind this analysis would be worth on your own numbers if every check fired, and until it runs nobody knows whether anything is wrong at all. 1 of the recommendations it can raise have no money formula, so they are not in that number.

Content

2 analyses
Do you have pages for what customers search for?The search demand exists whether or not you wrote a page for it. This looks at how much of it you are visible for at all.
Content · Built in, by us

No money ceiling on this one, because the recommendations it can raise have no deterministic formula on the data we hold. A made-up figure would be worse than none.

Does the traffic your content earns ever buy?Content can pull in readers who were never going to be customers. This checks whether the traffic you earn converts at all, and how it compares with pages you built to sell.
Content · Built in, by us
Could uncover up to ~19k kr/mo

That is a ceiling, not a promise: it sums what the recommendations behind this analysis would be worth on your own numbers if every check fired, and until it runs nobody knows whether anything is wrong at all. 1 of the recommendations it can raise have no money formula, so they are not in that number.

Shared by other operators

Definitions written outside our team, run by the same evaluator as ours, so a stranger's analysis can never compute a number its own way. Votes rank them and nothing else.

The shared catalogue could not be read just now, so only the built-in library is listed. Nothing below is affected by it.