Testing and refining
Find an attribution model that works for your business
Testing and refining
Attribution can be a complex, multi-dimensional challenge: customers engage over a period of time via multiple channels, on multiple devices, both online and offline.
We have the tools and specialist analysts to find the clarity you need, whatever your business goals, to make informed decisions on optimising your channel mix and spend.
Our scientific approach starts with identifying the right questions to ask, and then capturing the right data to provide the answers.
We then recommend changes, test the impact, re-run the models to identify further improvements, and constantly refine as more data becomes available.
If you are generating significant revenues from multiple countries, or have multiple brands, the Google Analytics 360 Suite can help analyse and compare extremely large data sets.
As official GA360 sales partners we can guide you through the data-driven attribution model in this platform, and how it integrates with DoubleClick and other products.
Questions about data-driven attribution?Ask one of our team
From our experience of working with some of the world’s leading brands, our starting point for building an attribution model is to ask what the model is for – the goals of the model.
Our analytics and data science specialists work with you to identify the questions that your attribution model should aim to answer.
These will vary for each company but will include questions such as:
Once we have identified the questions you need answering, we work with you to ensure that the right data is being captured for each touch point in the customer journey. This will include your analytics and advertising channels (e.g. AdWords, DoubleClick and Facebook), as well as offline data.
Depending on the volume of data, we will use a tool like Google BigQuery to analyse your data, looking for trends and building hypotheses.
We then make recommendations for changes to your marketing strategy, and test the impact and re-run the models to identify further improvements, constantly refining as more data becomes available.
“To run a successful campaign, you need a good relationship and a certain level of trust. In that respect we are in a really good place with Search Laboratory.”
Michal Chrusciel, MATCHESFASHION.COM
How well do you know your audiences?Speak to one of our analysts