7 Predictions for the analytics & martech space in 2023
Every year we try to predict what the future holds. We’ve done this last year and recently evaluated the quality of our predictions. With the new year kicking off it’s time to make predictions once again.
1) From walled gardens to black box fortresses
As privacy regulation is becoming more strict together with certain players (e.g: Apple) restricting data access and consumers becoming more privacy aware, advertising platforms are developing different methods and models to engage with users. Examples are the way Google models data in Google Ads or GA4. Where advertisers have a love-hate relationship with AI and machine learning, these trends towards ML automation and modelling will only increase in order to fill up the gaps created by privacy initiatives. Where marketers have been complaining about advertising platforms being black boxes (think of data-driven attribution in the past) this will reach new heights. It also implies that comparison between platforms will become harder and that attribution will happen per platform as the walled gardens become fortresses each running their own proprietary black box models.
2) Either you’re very good and know your customers, or you’ll be completely blind
Companies that aren’t investing in “identification rate” as a KPI and in building an actual relationship with their customers (see prediction 5) as a result of that will be completely blind going forward. Why? Because they’ll be at the mercy of the advertising platforms’ AI/ML models to bring in actual clients. Examples are Meta’s way of dealing with campaigns as well as how Google uses ML to build audiences. If you would go back 5 years and walk into a digital advertising agency you’d see analysts set up actual campaigns including targeting, geolocation etc. Today all these elements are automated and the platforms decide on where and when to spend money. In the context of 1st party data these platforms also request you to upload your existing customer base so they can look for more prospects. In short - if you don’t start building a relationship with your clients so you can start gathering your own data and insights, you’ll be at the mercy of the models living inside the black box fortresses.
3) Open source is back thanks to the Data Warehouse
In their attempt to secure data ownership we’ve seen many brands move into a data warehouse first approach and deploy additional open source tools in their own environments to avoid sharing data with third-party systems. Where years back environments would be on-prem and licence based and later on fully SaaS based we’re now seeing more organisations move into cloud and open source (or licence) based combinations. Examples are AWS, Azure or GCP based data warehouses topped off with tools such as Snowplow, Great Expectations, Metabase and Superset.
4) Snowplow will grow as a behavioral SDK
As a result of the previous point and the rise of the composable CDP we’re expecting Snowplow to grow as a behavioural data warehouse-first SDK. As many organisation are building their own version of a CDP most Reverse ETL solutions lack one major element (amongst others) - a behavioural SDK. Snowplow can fill this gap with grace.
5) BI & analytics tool capabilities will converge
Where BI tools such as Tableau and PowerBI used to be known for dashboarding and analytics tools such as Adobe, Mixpanel, Amplitude an Google Analytics used to be known for their analysis capabilities (e.g: sankey diagrams and funnel visualisations) more and more tools start to combine these capabilities. Most analytics products now have dashboarding capabilities just as most dashboarding tools include what used to be purely analytics visualisations. The rise of the data warehouse not only facilitates this trend but also creates new components or features for these existing tools. An example is the emergence of a new tool (maybe later feature?) category called metrics stores.
6) A set of new roles will emerge
We expect two main role to appear in organisations:
A. Privacy engineer
According to Wikipedia privacy engineering is “an emerging field of engineering which aims to provide methodologies, tools, and techniques to ensure systems provide acceptable levels of privacy”. In a world where privacy is becoming more important and the tension between legal and marketing/business/IT is caused by the mutual non-understanding of the respective field by one another and the implication when it comes to privacy a privacy engineer is here to build bridges between both worlds. The role also indicates the rise of a “privacy by design” philosophy that more and more brands are adopting rather than a “patching for privacy” philosophy.
B. Head of customer experience
This role might sound familiar as the title itself already exists. However, the role description and contents will change. Where today this role is typically part of a product or design team, it should actually be a company-wide transversal role focused on optimising every touchpoint a customer has with a brand. That goes from marketing to sales and support. This should also be the person who’s responsible (or at least involved) when it comes to customer data, the customer data infrastructure (including CDP ownership) and the experience delivery.
7) Ecosystem pitches and joint go-to-market strategies
Similar as in traditional media and marketing we predict that ecosystems will start to arise in the martech space. In traditional media brands will open up ecosystem RFPs where creative, online media and offline partners are invited to pitch together and propose a joint approach, strategy and pricing. This moves the complexity from the brand’s side to the agency side and forces agencies to ensure intercompatibility and operability before entering an RFP.
The way we see this happening in the martech space is that vendors not only will go to market together but will include service partners. This has as an objective to reduce the buyer and integrator friction as pricing should be simplified (not only the numerical aspect but also the procurement processes) as well adoption increased through smoother integration. Vendors in the CDP and analytics space will start to team up to provide a single - best-of-breed - offer that includes aligned pricing. Integration partners such as Human37 will be brought in from the start to ensure integration friction is reduced and the adoption accelerated.