Campaign measurement & reporting
Measuring and reporting campaign results
Measurement is where advertising either becomes accountable or stays a matter of opinion. This page explains what we track, how we report it, and which numbers we treat as evidence — including what cannot be measured cleanly.
The metric set fixed before launch
- 01 Impressions & reach Delivered volume and the number of different people or devices reached.
- 02 Frequency against target Average contacts per reached person, reported against the planned figure.
- 03 Clicks & click-through rate Interaction with the placement, shown next to the delivery it came from.
- 04 Video completion How much of a video placement was watched through, by format.
- 05 Cost per thousand, per click, per acquisition The spend metrics that let you compare channels on the same footing.
- 06 Conversions & cost per conversion Added for direct-response work, measured against your own outcome definition.
The list is agreed in writing before launch, so nobody argues about definitions later.
Metric set agreed before launch. Reported against target from the first delivery check.
The metrics we report
Every campaign gets a fixed metric set agreed before launch. That normally includes impressions, reach, frequency, clicks, click-through rate, video completion, and cost metrics such as cost per thousand impressions, cost per click and cost per acquisition.
For brand campaigns we add delivery against the frequency target, because repetition is what the plan was built around. For direct-response campaigns we add conversion volume and cost per conversion, measured against the outcome you define rather than the one that flatters the report.
Brand campaigns
Delivery checked against the frequency target, plus reach and total impressions for the period.
Direct response
Conversion volume and cost per conversion, with clicks and click-through rate as the early signal.
Definitions
Written down before launch, so a metric means the same thing in week one and in month six.
Reach and frequency explained
Zasięg is the number of different people or devices reached. Częstotliwość kontaktu is the average number of times each of them saw the ad. Both figures come from the same delivery data, and they always trade off against each other: at a fixed budget, buying more reach means buying less frequency.
Zasięg (reach)
Different people or devices, counted once
A person who sees the same ad four times is one reach and four contacts. That is why reach alone tells you how far the message travelled, not how often it landed.
Częstotliwość kontaktu
Average contacts per reached person
This is the average, so it hides the spread: some people see the ad once, others see it eight times. Where a platform exposes the distribution we show it next to the average.
The trade-off
One budget, two competing targets
We set both as targets in the plan and report actual delivery against them, because a campaign that misses its frequency target is not doing the job it was planned for. If you want the reasoning behind a specific split, the media planning process sets it out step by step.
See how reach and frequency targets are setEmisja reklamowa and delivery checks
Emisja reklamowa is the running of the advertising itself. Checking it means verifying that placements went live when scheduled, at the agreed weight, in the agreed formats.
We check delivery weekly in the first month and monthly afterwards. Discrepancies happen: a placement runs late, a format is substituted, a platform under-delivers. Catching these early is the difference between a corrected campaign and a report that explains a shortfall after the fact.
What a check covers
- Placements live on the scheduled date
- Agreed weight delivered for the period
- Formats as booked, or substitution flagged
When something is off
We raise it in the check itself, with the affected dates and placements listed, and propose either a make-good, a reallocation or a written note in the report.
Platform data versus independent measurement
Ad platforms report their own performance. Those numbers are useful, and they are also self-interested. Independent measurement — such as market-level ad monitoring that tracks online ad contacts and active brands in Poland — gives a separate view of how advertising behaves across the market. Where tracking on your own site or store exists, that is the strongest evidence, because it connects advertising to your own outcomes. We report all three where they are available, clearly labelled by source.
Platform-reported
Full delivery and cost detail, but graded by the party selling the inventory.
Independent market data
A market-level read on ad contacts and active brands, used as a second opinion.
Your own tracking
Site or store data tying spend to your outcomes — the strongest evidence available.
Attribution is never exact. We say which model a report uses, and we do not switch models mid-campaign.
Attribution and its limits
Attribution assigns credit for a conversion to the advertising that preceded it. It is a modelling exercise, not a reading. Last-click attribution overvalues search, because search often catches people who were already persuaded somewhere else. View-through and multi-touch models spread credit differently and produce different conclusions from the same data.
So we state which model a report uses, and we do not switch models mid-campaign to make a channel look better. If a model changes, the change is dated and the reason is written next to it, so you can compare like with like.
Last click
Simple to explain, generous to the final touchpoint, unkind to everything that built the demand earlier.
View-through & multi-touch
Spread credit across impressions and clicks, so upper-funnel work gets a share of the result it helped produce.
Reporting schedule and format
Reports follow a schedule agreed in the plan, usually monthly, with a weekly delivery check in the first month. Each report contains the agreed metrics, actual versus target, spend against budget, and a short written commentary on what changed and why. Raw figures are attached so they can be checked against platform dashboards. Reports are written in plain language, not a wall of charts with no conclusion.
Weekly through the first month, then monthly. Placements, weight and formats verified against the booking.
Agreed metrics with actual against target, spend against budget, and a short written commentary on what moved.
Raw figures attached, labelled by source, so any number in the summary can be traced back to where it came from.
Which model the numbers use, and whether it changed since the previous period.
What we cannot measure
Some things resist clean measurement. Brand recall, long-term preference and offline influence are hard to isolate, and any agency claiming to measure them precisely is overstating. Cross-device behaviour is only partly visible: a person who sees an ad on a phone and buys on a laptop is rarely a single clean record.
Where measurement is limited, we say so
We do not substitute a proxy for a real number and call it the real thing. Knowing which figures are solid and which are indicative is part of using a report properly — and it is the part that decides how much weight a decision can carry.
-
Brand recall and long-term preference: indicative at best, and usually better read through a market study than a campaign dashboard.
-
Offline influence: a shop visit or a phone call triggered by an online ad is often visible only in aggregate.
-
Cross-device paths: partly visible, and reported as such rather than merged into one confident figure.
Using reports to change the plan
A report is only useful if it changes something. If a channel is delivering below target on cost per acquisition for two consecutive periods, we propose cutting it or reallocating the budget. If a creative variant is clearly outperforming, we shift weight to it. Every change is recorded with the date and the reason, so the campaign's history stays readable. Plans that never change are usually plans nobody is reading.
The same discipline applies when a result is good. A channel that beats its target gets a written reason for the increase, not a quiet bump, so the next report can show whether the reason held.
Two periods below target
We propose a cut or a reallocation, with the numbers that triggered it.
One variant clearly ahead
We shift weight toward it and note the date, so the comparison stays honest.
Every change logged
Date and reason recorded in the report, so the campaign's history stays readable.
Questions we get about reporting
If your question is not here, the contact page reaches the team directly and we will answer it in plain language.
Contact the teamHow soon after launch do we get the first numbers?
The first delivery check lands in the first week after launch, because that is when late placements and format substitutions are still cheap to fix. The first full performance report follows the schedule agreed in the plan, usually at the end of the first month.
Can we compare your numbers with the platform dashboards?
Yes — the raw figures are attached to each report and labelled by source, so any number in the summary can be traced back to the dashboard or monitoring dataset it came from. Where a platform figure and an independent figure disagree, we show both rather than picking the better-looking one.
Why do reach and frequency targets matter so much?
Because they describe the job the campaign was bought to do. A brand plan built around częstotliwość kontaktu is not doing that job if delivery lands short of the target, even when impressions look healthy. We report the shortfall with the same prominence as an over-delivery.
What if a number in the report looks wrong to us?
Tell us, and we trace it. The source appendix exists exactly for that: a definition, a placement list and a raw export are usually enough to settle whether a figure is misread or genuinely off. If it is off, the correction is noted in the next report rather than quietly dropped.
Where this fits with the rest of the plan
Measurement only means something next to the plan it measures. If you want to see how channels and formats are chosen, or how a budget is split before any reporting starts, those pages explain the steps that come before this one.
Company registration
TELEKOMUNIKACJA AI Sp. z o.o.
- KRS
- 0001178761
- NIP
- 9542889925
- REGON
- 542055062
ul. Warszawska 40, lok. 2A, 40-008 Katowice, Poland
0622539213 · [email protected]
Monday to Friday, 9:00 AM – 6:00 PM
Metric set, reporting schedule and attribution model — agreed in writing before the first placement runs.