Google Meridian: Campaign effectiveness or cost escalations?

A week ago, Google made Meridian available to all brand marketers, ad agencies, and data scientists. Aimed to provide a 360-degree view of marketing activities across channels, this open-source marketing mix model (MMM) allows advertisers to make spending decisions based on data.

MMMs existed even for the offline media chanels. However, in today’s AI-dominated world, the conventional MMMs are seen ineffective in gauging the performance of digital media including search advertising and AI-powered campaigns. Meridian tries to help brand advertisers overcome these issues by incorporating advanced statistical models, such as Bayesian causal inference, to evaluate their marketing investments better.

Meridian’s ‘MMM data platform’ features metrics including impressions, clicks, and cost for Google media, in addition to latest parameters such as Google Query volume. These statistical data points assist advertisers in knowing exactly how paid search has been influencing sales. Meridian being an open-source framework, brand marketers can customise their code to suit their specific requirements.

Brand marketers can set key performance indicators (KPIs) specific to their business including sales, website visits, and conversions and scrutinise their campaign’s performance. The model also allows for including external aspects including pricing and promotions, to allow brand advertisers to carry out to holistically evaluate the effectiveness of their campaigns.

The traditional systems relied heavily on impressions. However, Meridian also includes tools to measure the aspects such as reach, frequency, and repeated exposure, providing a deeper insight into the impact of the ad campaign and the users’ behavioural patterns.

In Google Ads, incrementality experiments are tools that brand advertisers can use to understand the causal impact of their ads–whether they are driving the specific actions they care about at a certain point in time. Meridian now includes the results from incrementality experiments as well, allowing advertisers to compute their models with real business results. This improves the model’s ability to predict accurately, orienting the metrics more in line with business objectives.

From the brand advertisers’ perspective, Meridian helps them modify or tweak their budget allotments by comparing the effectiveness of their campaigns, including their offline performance, across different channels, in an integrated fashion. The model’s ability to include varying external factors such as seasonality of demand, economic conditions, major events impacting an industry’s performance (positively or adversely), etc, can further make their campaign evaluations more realistic.

Advertising agencies that manage large-scale, multi-channel campaigns, can leverage Meridian’s ability to integrate new and existing data sources and the flexibility it offers to tailor their models to specific needs of their clients for improved transparency and effectiveness. Meridian allows for adding partner agencies trained and certified in Meridian from Google’s ecosystem of partners. Using this feature, advertisers can gain access to specialised expertise to improve their campaign efficiency.

While Meridian offers several benefits to brands and agencies, there may be a few challenges on the way. For instance, the model’s Bayesian causal inference calls for knowledge of advanced statistical methods and data science expertise. Not all brands or ad agencies, especially from India’s MSME sector may have these capabilities available in-house, making it necessary to take extensive help of the certified agency partners from Google ecosystem, increasing the cost of the campaigns.

Similarly, while open-source nature of the model provides customisability, the small and mid-sized agencies in India will need to hire specialised coding talent or again, rely on the ecosystem partners of Google. This increased dependency on Google’s Meridian certified partners may create new revenue opportunities for those partner organisations, but will lead to cost escalation for ad agencies, and in turn, for brands.

Lastly, while Meridian’s cross-platform integration offers ability to brands to assess their campaigns across channels, agencies that have already invested in third party tools such as HubSpot, HootSuite, Adobe Analytics, SemRush, etc, will feel their investments are going waste. Alternatively, Google may come up with an option to allow for seamlessly using multiple tools while reducing the costs to users and protect their investments.

In a nutshell, Meridian presents a great opportunity to brands and agencies to improve the accuracy of their campaign performance management, but it also leads to fresh questions about return on assets (RoA) and cost escalations due to the need for specialised certified talent.