We are constantly innovating our marketing analytics platform and introducing new features and functionalities. This time we are proud to present six new features in our Augmented Analytics module,...
What is augmented analytics? As defined by Gartner, it is “the use of enabling technologies such as machine learning and AI to assist with data preparation, insight generation and insight explanation to augment how people explore and analyze data in analytics and BI platforms.” Or, to keep it simple – it’s a way artificial intelligence analyzes your data and delivers conclusions and suggestions on how you can improve your results.
Continuous improvement is one of the key methodologies that a company can use to gain a significant competitive advantage. And there’s no better way to improve than acting on data-driven insights. But getting more out of your data is not that easy. You can easily get lost in the abundance of information. So many tiny, but potentially impactful opportunities for improvement can easily be overlooked. Also, a tiny human error can cost your company lots of money and bring you, personally, a lot of frustration.
With all these situations in mind, we have developed a set of intelligent features that can help you to ‘find a needle in the haystack’ and successfully identify and eliminate even the smallest errors. We are proud to present the six new features of our Augmented Analytics module: Anomaly Detection, Trend Detection, Segment Analysis, Budget Shift Suggestions, Forecasting, and Channel Mix Optimization.
The Anomaly Detection feature collates historical data on many different metrics, such as impressions, clicks or CPM, and uses it to learn and set future expectations for these metrics. If there is an anomaly detected, either as an unexpected drop or a spike, this outlier is flagged, and a graphical chart is automatically created to visualize it.
The projections are calculated using a rolling standard deviation method, detecting, as outliers, those data points that stand outside a certain standard deviation of the time series. The size of the standard deviation is adjusted daily, with the addition of each new data point, which constantly improves the quality of the anomaly detection function.
The main benefit of this feature is the early detection of outliers across many different metrics, often caused by the smallest human errors, such as mistyped CPM budget values or missing conversion pixel tracking. This enables data analysts or performance marketers to identify potential problems as early as possible, before they become a major issue. Agencies can use this feature to provide immediate insight to their clients and optimize use of their budgets, while advertisers can avoid otherwise unnoticed budget drain or performance issues.
For an advanced way of detecting trends within mountains of marketing data, the Trend Detection feature collates historical data on important metrics, which it then intelligently normalizes to allow for weekly and yearly seasonality, reduces noise within the data, and determines the overall trend across time. The detected trends are then scored to determine their relevance, and the user is informed through an automatic creation of a graphical chart showing any significant trends which may affect campaign performance, or overall business results.
This feature enables organizations to easily and quickly identify emerging trends which may be of importance to them, such as major changes in the market or unforeseen events. It reduces the time otherwise spent manually researching those trends, and highlights trends which may actually have been missed. Agencies can demonstrate added value to clients by identifying trends that can significantly improve the client’s business, while brands and advertisers can make better business decisions and proactively optimize campaigns for future success.
The Segment Analysis feature captures all segmented data being fetched from relevant data streams. It analyses all breakdowns of segments, such as age, geography or device, across any channel, which could be potentially hundreds of combinations. It then clusters similar segments, with the same target metric, and analyses the relative performances of each segment within a cluster. When an out-performing sector is identified within a cluster, a visualization is created to inform and visualize this segment and its performance.
While the audience segment usually has been defined before the campaign launch, this feature provides a more granular analysis into the most effective segments emerging as that campaign runs. It enables the initial targeting to be refined and completely re-defined if necessary, based upon insights learned as the campaign progresses. With this feature, brands and advertisers can adjust campaign budgets and activities to be focused on more specific market segments, likely to produce better results.
The Budget Shift Suggestions feature classifies campaigns, and clusters similar ones together. It then automatically identifies the target metric for each cluster. By analyzing the relative performances of each campaign within a cluster, it identifies those which are performing relatively well or badly. Finally, it determines whether those campaigns are being restricted by budget, and provides visual suggestions of where increasing that budget would actually improve performance, and where budget is being wasted, and should be reduced.
The suggestions generated by this feature are designed to be very actionable and can be immediately implemented. Budget Shift Suggestions provides an organization with insights on spend versus performance for each campaign, allowing them to quickly reallocate budgets to achieve the optimum performance across all campaigns and improve ROI. For agencies, this feature can contribute significantly to the added value they deliver to the clients, as it brings a highly optimized way of managing a large number of campaigns.
The Forecasting feature informs you of how much you have already spent in a certain period, and what your predicted spend will be by the end of the period. The prediction is based upon historical spend data, with typical seasonal variations applied. It provides values for those figures, as well as a graphical view showing the actual daily spend so far and the predicted daily spend for the future. You can apply the forecast feature to all digital marketing activities, or filter on specific entities, such as platforms, accounts or campaigns, or a mixture of any.
The forecasting feature provides a true picture of spend during a defined period, and enables you to stay on track, by providing accurate and timely suggestions for changes in spending, which you can then act upon to ensure that spending is ultimately not over or under budget. It allows marketers to continually monitor and fine-tune their spending in order to achieve the optimum utilization of the allocated budget, and reduces the amount of time they would normally use for monitoring and adjustment of campaigns.
The Channel Mix Optimization feature determines what mix of spend across marketing channels will result in the maximum revenue for a given budget. The feature takes into account data on past costs from each marketing channel, and revenues gained from campaigns on these channels. It then builds a historical picture of ROI across those channels, and by using a statistical regression model per channel it learns how spend on each channel has affected revenues over the same time period. The longer the historical time period, the more accurate the results.
There is a graphical representation of the current and optimal spend for each channel, as a way to easily understand how to reallocate the budgets. It provides the user with a clear, graphical comparison of current budget allocation for each channel or a set of channels, and how that allocation should be changed, allowing you to make fast decisions to continually achieve the best performance.
With all of these features, marketers and agencies can continually assess the performance of their digital campaigns, and fine-tune budget allocations and campaign parameters to achieve the maximum marketing ROI. In a time where budgets are getting smaller and there’s constant pressure to grow further, this can be a crucial argument in establishing marketers as the key growth drivers in companies of all sizes.