Revenue without a source
42.6% of revenue has no UTM source (18 400 € of 43 200 €)
- Home
- Insights
Insights: what we know, what we assume, what comes next
KO goes through your project data on its own — ads, website, leads, deals and the plan — and writes a card when something needs a decision. The card keeps facts, likely causes, the forecast and the next step apart. Here is every kind of insight KO finds.
A number is not an answer yet
A report shows how much. The real question is usually different: what went wrong, why, and what to do about it. People look for the answer by hand, with exports and spreadsheets, and often find it after the mistake has already cost money.
- A dashboard shows a number An insight says what is wrong with it, why, what to do, and what happens if you do nothing.
- Someone reads the numbers when there is time KO goes through them on its own and raises only what needs a decision.
- Data is taken on trust KO first checks whether the data can be trusted. Broken tagging is an insight too.
- The number dropped, so something must change Not always. If holidays or the season moved the total, KO says so: an explanation, not a task. What works is marked “better leave it as is”.
Insights by the question they answer
- Role in the question:
- Answers The finding answers the question directly: where, how much, what to do.
- Precondition Until this is fixed, the answer can't be trusted: the check says data or tagging is missing.
- Guard Does not answer the question itself but stops a wrong move: shows where the obvious answer would be wrong.
- Result type:
- Broken measurement
- Confirmed
- Likely
- Forecast
- Opportunity
- Observation
- Healthy
The cards and messages on this page are examples. The numbers are made up: in the product, they are the numbers of your project.
No insight matches this search.
Where to move the budget
- Losses by campaign Answers Likely Shows the campaign with noticeably more refusals among its leads.
- Conversion by channel Answers Likely Shows the channel whose leads reach a deal noticeably less often.
- Ads stop bringing people in the middle of the day Answers Likely The “brand_search” campaign (google): on 5 of 7 days the impressions stop around 14:00, while the rest of the site gets 41% of its visits after that hour
- A keyword brings leads much more or much less expensive than the others Answers LikelyOpportunity Names keywords inside a campaign to lower the bid on, and keywords to raise it for.
- Overspend against the norm Answers Likely Names campaigns where a result costs more than the project norm: candidates to take money from.
- Best channel on small volume Answers Opportunity Answers 'let's pour money into the best channel': shows how much it adds even if doubled.
- Sales speed by source Answers Opportunity Shows which source brings quick sales and which brings slow or empty leads.
- A new campaign ran out of steam Answers Likely The “Autumn — Lead” campaign brings 46% fewer requests per click than in the first 21 days after the launch
- Channel pause and other leads Answers LikelyObservation Checks a budget move already made: what happened to the other requests after a channel was paused or its budget changed.
- Ads eat the value of a new sale Answers Likely Ads take 70% of new-sale revenue, against 40% in 06.2026: a new sale went from 100 € to 182 € in ads (+82%), and the average order value grew 5%
- Clutter in the source field Precondition Broken measurement If the source field holds a contact method, channels are compared against messengers, not each other.
- Leads without a source Precondition Broken measurement Leads without a source fall into no channel, and the channel comparison is skewed.
- Spend linked to results Precondition Broken measurementForecast If spend is not linked to visits and leads, campaigns can't be compared by result.
- Depth of ad tagging Precondition Broken measurement Without a campaign tag, money can move only between channels, not between campaigns.
- Revenue without a source Precondition Broken measurement If most revenue has no source, channel returns are computed from a minority of deals.
- Sales without a source Precondition Broken measurement If many sales have no source, the 'best' channel may just be the best tagged.
- Budget by region Precondition Broken measurement If regions are not visible in campaigns, moving money between regions is blind.
- Attribution window vs sales cycle Guard Broken measurement If deals close slower than the report window, a slow channel looks worse than it is.
- Sales hit the managers’ workload Guard Likely If managers can't handle the flow, more leads will not bring more sales.
- The share of non-human visits has grown Guard Broken measurementLikely Cheap visits may not come from people: the channel looks efficient, but there is nobody to buy.
- Leads depend on one channel Guard Opportunity Before adding money to a strong channel: the project already depends on it, and the risk grows.
- Current channels will not reach the goal Guard Forecast Shows that moving money between current channels will not close the goal: a new channel is needed.
- A source sent a spike of leads, but few of them are target leads Guard Likely "partners" sent 48 leads in a week against the usual 20, and 12% became target leads — usually 35%
Why leads or sales dropped
- Drop in tagged leads Answers Broken measurement Attributed leads fell from 82.4% to 51.3% against the baseline period (−31.1 pp)
- A source converts worse week after week Answers Likely “Google”: conversion to sale 9.1% → 4.2% over 8 weeks with the same lead flow
- More leads are lost at a late stage Answers Likely After the “Proposal sent” stage we lose 66% of leads that got there, against 39% four weeks earlier
- A landing page asks for an A/B test Answers Opportunity Of visits to the page /services, 0% leave a request, against 1.1% on the rest of the site: the page is 100% behind
- Data gone quiet: a stream stopped Answers Broken measurement The “Contact form submissions” stream has brought no data since 2026-09-14. That is 14 days in a row, while it normally pauses for about 2 days.
- Pace against the previous period Answers LikelyForecastHealthy “Leads”: 142 over the last 30 days against 205 over the previous 30 days, down 31% per working day
- Why a source lost its leads Answers LikelyBroken measurement Leads from “google” seem to lose their source tag from the week of 2026-09-07: 15 per week counted instead of 42
- A conversion gain did not hold Answers Likely “google”: conversion to sale went up from 6.7% to 31.1% and came back to 10.9%
- A new campaign ran out of steam Answers Likely The “Autumn — Lead” campaign brings 46% fewer requests per click than in the first 21 days after the launch
- Mix or behavior Answers ObservationHealthyForecastBroken measurement Lead conversion for the last 30 days: 4.2% → 3.1% (−1.1 pp). The mix by “Source” changed, not the result inside the groups
- The share of non-human visits has grown Guard Broken measurementLikely If visits grew but requests did not, part of the growth may not come from people.
Can you trust the numbers
- Drop in tagged leads Answers Broken measurement Attributed leads fell from 82.4% to 51.3% against the baseline period (−31.1 pp)
- Lost leads without a reason Answers Broken measurement 41.2% of closed deals have no loss reason over the last 30 days (7 of 17)
- CRM stage mapping Answers Broken measurementOpportunity 12 active leads sit on 2 funnel stages with no type
- Leads without an assignee Answers Broken measurement 3 of 58 leads in the CRM funnel (5.2%) have no assignee over the last 30 days
- Clutter in the source field Answers Broken measurement 14.2% of leads with a filled-in source (29 of 204) have something other than a source there
- Leads without a source Answers Broken measurement 46.8% of leads have no known channel over the last 30 days (73 of 156)
- Spend linked to results Answers Broken measurementForecast 4 820 € of 15 360 € in ad spend (31.4%) links to neither leads nor visits
- Depth of ad tagging Answers Broken measurement 14 of 52 paid leads (26.9%) have no UTM campaign over the last 30 days
- Attribution window vs sales cycle Answers Broken measurement Only 37.5% of deals fit inside a 30-day window (9 of 24). The median sales cycle is 41 days
- Revenue without a source Answers Broken measurement 42.6% of revenue has no UTM source (18 400 € of 43 200 €)
- Sales without a source Answers Broken measurement 40% of sales have no UTM attribution over the last 30 days (4 of 10 won deals)
- The share of non-human visits has grown Answers Broken measurementLikely Among visits without a tag, 48% in recent weeks are non-human, against 21% before
- Data gone quiet: a stream stopped Answers Broken measurement The “Contact form submissions” stream has brought no data since 2026-09-14. That is 14 days in a row, while it normally pauses for about 2 days.
Will we reach the goal
- Current channels will not reach the goal Answers Forecast Sales · Monthly plan: forecast by the end of the month is 19, goal 29, gap 10
- Plan pace Answers ForecastBroken measurement “Leads”: at the current pace the month ends at 161 against a plan of 220 (73%)
- Pace against the previous period Answers LikelyForecastHealthy “Leads”: 142 over the last 30 days against 205 over the previous 30 days, down 31% per working day
- Leads without sales Guard Confirmed The lead plan is met, but there are no sales: a volume goal misleads.
Where leads get lost after the request
- Leads without activity Answers Confirmed 14 active leads (23%) with no activity in the CRM for too long
- Funnel bottleneck Answers Likely 14 of 31 active leads (45.2%) are held on stage “Proposal sent” longer than usual
- Leads closed without a sale Answers Likely 71.4% of leads over the last 30 days were closed without a sale (60 of 84). This is above the project's usual 58.2%
- Leads stuck at a stage Answers Likely 12 leads with no stage change for a long time on intermediate funnel stages
- Deals take longer to close Answers Likely From request to sale it now takes 32.3 days instead of 18.9, 71% longer
- Conversion by manager Answers Likely “Mikko Laine” converts at 18.2% against 46.2% for “Anna Virtanen” over the last 30 days
- No sales to new customers Answers Confirmed No deals among 12 resolved leads that have had time to go through the usual sales cycle
- Leads without sales Answers Confirmed 34 leads over the last 30 days. The plan is “met”, but sales are 0
- Sales hit the managers’ workload Answers Likely In busy weeks the managers convert 5.8% against 11.9% in usual weeks
- More leads are lost at a late stage Answers Likely After the “Proposal sent” stage we lose 66% of leads that got there, against 39% four weeks earlier
- A refusal reason is concentrated in one source or with one manager Answers Likely “Not a fit” is the reason for 35% of “facebook-ads” refusals, against 22% in other sources
- Leads that come in off hours buy less often Answers Likely Leads that came in off hours buy in 2.1% of cases, in working hours in 4.0%
- Price barely causes refusals — a higher order value is worth a test Answers Opportunity Of 64 refusals with a named reason, 2 (3%) are because of price, and 31% of closed deals are won against 27% for the year
- Low-cost first product: is there one, and does it lead to a full purchase Answers Opportunity Of 84 buyers who started with a purchase under 294 €, 5 reached the full one (6%)
- Lost leads without a reason Precondition Broken measurement 41.2% of closed deals have no loss reason over the last 30 days (7 of 17)
- CRM stage mapping Precondition Broken measurementOpportunity 12 active leads sit on 2 funnel stages with no type
- Leads without an assignee Precondition Broken measurement 3 of 58 leads in the CRM funnel (5.2%) have no assignee over the last 30 days
What worked and is worth keeping
- Channel pause and other leads Answers LikelyObservation “facebook-ads”: paused from the week of 24.08, spend €1,450 → €0 per week. Other leads in the same weeks: 31 → 19 per week
- What works Answers Healthy Improved by 32% for channel “google-ads”: conversion to sale rose from 18.5% to 24.4% against the previous 30 days
- Attribution window vs sales cycle Guard Broken measurement If deals have not had time to close yet, it is too early to judge a change.
- Mix or behavior Guard ObservationHealthyForecastBroken measurement Shows whether the work improved or the mix of traffic simply changed.
Data quality
Checks whether CRM and tagging data can be trusted: leads with no assignee, no activity, no loss reason or no source, unmapped funnel stages, clutter in the source field, drops in tagging. Until these problems are fixed, business conclusions rest on incomplete data.
- Drop in tagged leads Broken measurement Attributed leads fell from 82.4% to 51.3% against the baseline period (−31.1 pp)
- Lost leads without a reason Broken measurement 41.2% of closed deals have no loss reason over the last 30 days (7 of 17)
- CRM stage mapping Broken measurementOpportunity 12 active leads sit on 2 funnel stages with no type
- Leads without an assignee Broken measurement 3 of 58 leads in the CRM funnel (5.2%) have no assignee over the last 30 days
- Leads without activity Confirmed 14 active leads (23%) with no activity in the CRM for too long
- Clutter in the source field Broken measurement 14.2% of leads with a filled-in source (29 of 204) have something other than a source there
- Leads without a source Broken measurement 46.8% of leads have no known channel over the last 30 days (73 of 156)
Measurement problems
Looks for places where measuring ad return is broken: spend not linked to results, sales and revenue with no source, tagging only down to the channel, an analysis window shorter than the sales cycle. These cards come first, because without them channel comparison cannot be trusted.
- Spend linked to results Broken measurementForecast 4 820 € of 15 360 € in ad spend (31.4%) links to neither leads nor visits
- Depth of ad tagging Broken measurement 14 of 52 paid leads (26.9%) have no UTM campaign over the last 30 days
- Attribution window vs sales cycle Broken measurement Only 37.5% of deals fit inside a 30-day window (9 of 24). The median sales cycle is 41 days
- Revenue without a source Broken measurement 42.6% of revenue has no UTM source (18 400 € of 43 200 €)
- Sales without a source Broken measurement 40% of sales have no UTM attribution over the last 30 days (4 of 10 won deals)
Funnel analysis
Problems in the sales funnel: leads stuck at stages, more losses than usual, no sales to new customers, many leads but zero sales, a gap between managers. Until the funnel has enough history, conclusions are shown as a forecast.
- Funnel bottleneck Likely 14 of 31 active leads (45.2%) are held on stage “Proposal sent” longer than usual
- Leads closed without a sale Likely 71.4% of leads over the last 30 days were closed without a sale (60 of 84). This is above the project's usual 58.2%
- Leads stuck at a stage Likely 12 leads with no stage change for a long time on intermediate funnel stages
- Deals take longer to close Likely From request to sale it now takes 32.3 days instead of 18.9, 71% longer
- Conversion by manager Likely “Mikko Laine” converts at 18.2% against 46.2% for “Anna Virtanen” over the last 30 days
- No sales to new customers Confirmed No deals among 12 resolved leads that have had time to go through the usual sales cycle
- Leads without sales Confirmed 34 leads over the last 30 days. The plan is “met”, but sales are 0
- A source converts worse week after week Likely “Google”: conversion to sale 9.1% → 4.2% over 8 weeks with the same lead flow
- Sales hit the managers’ workload Likely In busy weeks the managers convert 5.8% against 11.9% in usual weeks
- More leads are lost at a late stage Likely After the “Proposal sent” stage we lose 66% of leads that got there, against 39% four weeks earlier
- A refusal reason is concentrated in one source or with one manager Likely “Not a fit” is the reason for 35% of “facebook-ads” refusals, against 22% in other sources
- Leads that come in off hours buy less often Likely Leads that came in off hours buy in 2.1% of cases, in working hours in 4.0%
Traffic analysis
Compares channels and campaigns by conversion, losses and cost per result. Checks whether you can see which regions the budget goes to. Answers the question of where to move money.
- The share of non-human visits has grown Broken measurementLikely Among visits without a tag, 48% in recent weeks are non-human, against 21% before
- Losses by campaign Likely Campaign “Home cleaning / Espoo / Meta” loses 58.3% of leads, against 24.6% on average for the other compared campaigns over the last 30 days
- Conversion by channel Likely Channel “linkedin” converts at 12.5% against 50% for “google-ads” over the last 30 days
- Budget by region Broken measurement The region shows in the name of 2 of 7 campaigns with spend, which is 23.9% of the budget over the last 30 days
- Ads stop bringing people in the middle of the day Likely The “brand_search” campaign (google): on 5 of 7 days the impressions stop around 14:00, while the rest of the site gets 41% of its visits after that hour
- A keyword brings leads much more or much less expensive than the others LikelyOpportunity The keyword “coffee machine repair” in the “search_main” campaign (google) took 38% of the campaign spend and brought 1 lead of 14 — about 5 were expected for its share of spend
- Overspend against the norm Likely Campaigns spent 1 840 € above the project norm over the last 30 days (11.6% of paid spend)
Opportunities
Looks for unused potential: the best channel with small volume, sources with fast or slow sales.
- Best channel on small volume Opportunity The best converting channel “linkedin” (45%) covers only 6.6% of the pool (14 of 212 leads) over the last 30 days
- Sales speed by source Opportunity Leads turn into money faster: source “google” has 38.5% fast sales against 21.4% for the rest, among leads from 15.02.2026 to 14.08.2026
- Leads depend on one channel Opportunity google, paid — 82% of leads with a known channel (205 of 250) over the last 90 days
- Current channels will not reach the goal Forecast Sales · Monthly plan: forecast by the end of the month is 19, goal 29, gap 10
- A landing page asks for an A/B test Opportunity Of visits to the page /services, 0% leave a request, against 1.1% on the rest of the site: the page is 100% behind
- Price barely causes refusals — a higher order value is worth a test Opportunity Of 64 refusals with a named reason, 2 (3%) are because of price, and 31% of closed deals are won against 27% for the year
- Low-cost first product: is there one, and does it lead to a full purchase Opportunity Of 84 buyers who started with a purchase under 294 €, 5 reached the full one (6%)
Plan
Compares the forecast at the current pace with the project plan from the Planning section: falling behind, going over target, overspending the budget by the end of the period.
Anomalies
Catches a broken data stream (a source went silent or its volume dropped sharply), a change in the pace of leads, sales and revenue against a baseline period, and a sharp drop in leads of a paid source with its cause: traffic, visit quality or tagging.
- Data gone quiet: a stream stopped Broken measurement The “Contact form submissions” stream has brought no data since 2026-09-14. That is 14 days in a row, while it normally pauses for about 2 days.
- Pace against the previous period LikelyForecastHealthy “Leads”: 142 over the last 30 days against 205 over the previous 30 days, down 31% per working day
- Why a source lost its leads LikelyBroken measurement Leads from “google” seem to lose their source tag from the week of 2026-09-07: 15 per week counted instead of 42
- A source sent a spike of leads, but few of them are target leads Likely "partners" sent 48 leads in a week against the usual 20, and 12% became target leads — usually 35%
- A conversion gain did not hold Likely “google”: conversion to sale went up from 6.7% to 31.1% and came back to 10.9%
- A new campaign ran out of steam Likely The “Autumn — Lead” campaign brings 46% fewer requests per click than in the first 21 days after the launch
- Mix or behavior ObservationHealthyForecastBroken measurement Lead conversion for the last 30 days: 4.2% → 3.1% (−1.1 pp). The mix by “Source” changed, not the result inside the groups
- Channel pause and other leads LikelyObservation “facebook-ads”: paused from the week of 24.08, spend €1,450 → €0 per week. Other leads in the same weeks: 31 → 19 per week
Money without results
Looks for places where ads spend money and bring no result: a channel or campaign pays for weeks, and there are no leads. Counts the loss in money and tells a weak channel from broken lead tracking.
What works
One card about what has clearly improved. It helps keep what already works and repeat it in similar areas.
What is inside an insight
The card is a short conclusion. A click opens the breakdown: what the conclusion stands on and how far you can trust it.
- What we know
Facts with numbers and their source: what changed, by how much, compared with what.
- Where exactly
The channels, campaigns, managers or stages where the change is strongest, with a link to the report.
- Cross-check
The conclusion is checked against a second signal: confirmed, not confirmed or cannot be checked. Plus what else changed on the same days.
- We assume
Likely causes. A hypothesis is labelled as one and never passed off as a fact.
- We forecast
Where the period ends if nothing changes.
- What to do
The step, what it gives, how to check the result, and the cost of doing nothing.
- For
Owner, analyst or agency client: each role sees the cards in its own order.
- More than one condition behind a card
To raise one card, KO goes through the slices of your project, including the fields of your CRM, and compares them with the norm of your own project, adjusted for season, holidays and working days. The conclusion is checked against a second signal. When KO finds the root cause, its effects do not show up as separate cards: you see the cause, not ten symptoms.
Demo
Imagine your project is already in KO
The demo is the project of a made-up company, with ads, website and CRM already connected. Open “What to do now” to see what your insights would look like: each with its breakdown, result type and who it is for.
See it live
No sign-up, no payment
See where your revenue comes from
See what to cut, fix, and scale — using your real data.
You'll meet the founders, not a sales rep — we built KO after years of running growth on broken numbers.