What do the figures show before and after a change? We compare up to five metrics and review other influences that limit the conclusions.
Your enquiry is non-binding. An order is created only when you accept our quotation .
Image 1 of 4
Orders have increased since the new product page. A campaign is also running. We compare available data before and after the change and assess which conclusions it supports.
Impact analysis examines one documented change, up to five agreed metrics and two comparison periods of no more than eight weeks each. It uses one analytics property or agreed export. A usable baseline and change date must be available.
We first check whether the same events and periods are being compared. We also record measurement changes or a campaign running at the same time. Temporal coincidence is not hastily presented as the effect of the change.
The report presents the comparison, data quality and open questions, leading to a recommendation for the next investigation. A controlled test can provide a different basis from time comparison alone; its planning and technical setup are excluded. If suitable data is missing, no impact result is invented.
The order rate rises after redesign, while a discount campaign also ran. We compare the agreed periods and document these influences before attributing the difference to design.
The team treats more clicks as success while received enquiries stay unchanged. Before analysis, we define up to five metrics addressing the question examined.
The figures show a difference but do not yet explain its cause. The report identifies the open question and recommends a next test to investigate it more specifically.
We record the change, date and metrics to review. We then choose two comparison periods with measurement definitions and conditions that align as closely as possible.
We assess whether data can answer the question and record identifiable confounding factors. Gaps remain visible; we do not fill them with apparently precise figures.
The report shows observed differences and explains what cannot be inferred. It recommends a next question to test. A/B-test setup, tracking repair and implementation are excluded.
Include the comparison and its limitations in planning. Rising click counts justify a different next step from more completed orders. Where several things changed together, the suspected cause needs further investigation.
If the measurement baseline itself is unclear, an analytics data check may be useful first. A quarterly review considers an existing baseline in a recurring project context. Where the same period is already examined there, included checks are not charged again as impact analysis.
Which decision depends on the figures? Note the metric, analytics tool and known changes during the period. We clarify these together after receiving your enquiry.
You do not need to provide credentials or full exports yet. We first agree which data is needed and how it can be supplied.
What is included, what do you need and what happens after your enquiry?
The process in detail →No. A before-and-after comparison alone does not prove causality. Other influences may occur simultaneously. The report makes known confounding factors and data gaps visible rather than attributing every change to the latest action.
A documented change with its date, a usable baseline and comparable data for two agreed periods are needed. Describe known campaigns, measurement gaps or further changes. Credentials must not go in the enquiry form.
A quarterly review may already cover the same change and period. In that case, scope is reconciled beforehand. Impact analysis is intended for a separately defined investigation, not charging twice for the same comparison.
The basis is one analytics property or agreed data export. Describe your tool, available metrics and documented change in the enquiry. Before commissioning, we clarify whether the data can be analysed appropriately. A new PostHog integration is not automatically set up.
Two comparison periods of no more than eight weeks each are included. Suitable periods depend on the documented change and data baseline. Arbitrarily selecting periods because they show a favourable difference would not provide reliable evidence.
No. A recommendation for the next test is part of the report; technical implementation is not. The package includes neither an experimentation platform nor tracking repair. Actual testing requires separate clarification of the question, technical requirements and scope.
Analysis limitations are documented. Missing values are not guessed to report success anyway. A sensible next step may be clarifying measurement or making a suitable comparison later rather than deciding on a change without evidence.
The data baseline, comparison periods and work window are confirmed before commissioning.
A usable baseline and documented change date Comparable data for the two agreed periods Known additional changes, campaigns or measurement gaps described where available Credentials must not be included in the enquiry. Details are agreed only after written commissioning.
One written follow-up round is included. The service scope lists the analyses, designs or technical changes covered by the package. We agree additional work with you before it begins.
How much time does your monthly report take? Include checking, setup and tool costs, and clarify conflicting figures before the next dashboard.
Read the guide ↗
Which events does your tracking actually count? We check one analytics property for gaps and conflicting figures.
What has changed since the initial diagnosis? We review agreed metrics, implemented changes and the next tasks.
