Benchmark dashboards for ad campaign optimization
We developed benchmark dashboards solutions for a digital media company to monitor ad campaigns and forecast their results, helping them improve benchmark achievement by 8x.
Context
The customer is a leading digital media company with over 1.5m subscribers and 50+m social media followers worldwide. Their content marketing channels are numerous and include mobile apps, podcasts, social media, landing and brand pages. The customer uses this vast marketing platform to promote their clients’ brands via organic and paid promotion methods and commits to delivering certain results.
Given the number of clients, content channels, and types and tiers of promotion packages, the company struggled to monitor ad campaign progress. They also had trouble achieving the benchmarks because they were static and didn’t take into account various factors. Due to this, static tables often misled client brands by creating unrealistic expectations. What is more, static tables didn’t allow sales teams to accurately predict the results of ad campaigns and meet the allocated budget, so only 62% of clients achieved the set benchmarks in the end.
On top of that, the customer would constantly update or extend their content channels, introduce new content types, and launch new target audience engagement and conversion methods. This meant they would need to not only track new metrics but also compare them to historical data.
Faced with these challenges, the customer decided to revamp their data analysis and visualization processes. The company chose Itransition as a consulting partner due to previous successful collaborations with us and our expertise in providing big data services to media and entertainment companies.
Solution
Itransition developed several data-driven analytics optimization solutions for monitoring ad campaigns and forecasting their results using dynamic benchmarks.
Forecasting solution
The forecasting solution our team built is a system of consolidated dashboards that use historical data. To help sales managers make accurate forecasts, we created a table with reference benchmarks and an algorithm for calculating multifactor benchmarks for six filters: Impressions, CTR, Channel, Region, Product, and Duration.
We also developed campaign model visualizations to better demonstrate the number of impressions across channels over time with CTR results. Campaign models can visualize different types of changes in linear, logarithmic, parabolic, and mixed progressions.
Analyzing the simulated models, the production team can create content for each channel based on specific user expectations, while the media team can monitor the overall performance as well as the performance across each channel.
Relying on clearly visualized multifactor dynamic benchmarks, the customer’s sales team can simulate promotion campaigns, calculate divergences from established benchmarks, make reliable forecasts, form realistic expectations, and recommend ad campaign changes.
Monitoring solution
The Itransition team also developed a monitoring solution called Campaign monitor for tracking ad impressions by choosing different filters. The solution is aimed at helping digital strategists and media managers to gauge campaign effectiveness and tweak advertising strategy accordingly.
We also equipped the monitoring solution with a recommended activity modeling feature. The digital strategist chooses a priority and a desirable benchmark and receives the list of recommended activities for reaching the desired goals in each channel. The feature is aimed at preventing budget overspending and enabling real-time ad campaign monitoring.
Using the solution, the customer’s media team can also perform plan-fact analysis, compare the planned and actual metrics, and track advertising plan completion, while analysts can monitor area charts and build forecast graphs.
Technologies
We build the campaign optimization solution based on Tableau to allow users to not only view data but also work with it via the solution’s interface. We also used a Tableau extension Data Writer to transmit data from parameter values selected by the user and allow them to connect their data sources and add Tableau parameters to the solution’s database fields.
Our team also developed a Python algorithm that communicates with the database and enables the calculation of multifactor correction coefficients. We also created a decision-making algorithm that combines sets of parameters and historical, actual and predicted data.
This way, the delivered solution encompasses management tools like dashboards, visual analysis like comparison charts, and the Python database algorithm as well as the Tableau extension performing mathematical calculations.
Process
The Itransition team began by developing the future solution’s PoC and tested it with historical data for the customer’s existing products and the latest quarterly data for new products. Next, we built the analytical model and performed tests with synthetic data to discover what parameters were missing and supplemented the initial dataset with information from across the customer’s departments.
Having improved the solution in line with customer feedback, we created the second prototype, tested it, and got stakeholders’ approval after several refinements. After final successful testing, we created detailed documentation for the solution, describing the updated scripts and how to use them for training sales, product and media teams.
Results
We developed a set of analytics optimization solutions equipped with dynamic dashboards and customizable campaign-specific benchmarks for a leading digital media company. Our customer can apply it to make realistic forecasts of the results of marketing campaigns, get real-time accurate predictions, and monitor advertising campaign performance through user-friendly dashboards.
Having implemented our analytics solution, the customer managed to improve benchmark achievement by 8x and reduce overspending by 7x, which resulted in a noticeable customer satisfaction level improvement.
