In the last several years, the analytics space has changed dramatically. Until recently, the challenge was combining and normalizing different data sources so that the result would be accurate and accessible. These days, aggregation is par for the course, with the most significant value seen (literally) in the form of visualization; the goal being to make data (already aggregated) more approachable and more easily digestible. As the saying goes, “A picture is worth a thousand words”.
Now, market leaders are beginning to move beyond visualization and push for more efficient, scalable, and insightful forms of data interpretation. Bridging the gap between visualization and interpretation requires a giant leap forward. To do this, BI solutions will for the first time be asked to actually make sense of data — not collect it, not transform it, but understand it.
While automation has already transformed monitoring into a “must-have” solution for DevOps teams, business teams have been left behind.
This is because available solutions fail to properly map the data relationships between the different metrics and dimensions they track. As a result, they cannot correct for context — weeding out irrelevant issues, weighing impact, or bundling connected developments (noise reduction). And because incumbent solutions bear no real understanding of how the data ties back into the business, they cannot translate mathematical observations into concrete operational problems and opportunities.
With the business side still relying on dashboards, manual data review, cobble-built in-house solutions, and simplistic rule-based alerts, they face a seriously uphill battle every day. A fact that comes with a cost — leading to significant time and revenue losses.
Now that’s changing with a new generation of business solutions that aim higher; that aim to deliver automated interpretation.
Interpretation is complex and messy work that relies as much on inductive reasoning as deductive reasoning — which is why it’s been left to humans until now. Opening this type of work to machines — that don’t get tired, don’t make mistakes, aren’t limited by scale, and can work around the clock — will unleash untold business value.
This is where an Israeli startup called oolo says it can help. Designed to help businesses more rapidly and effectively operationalize their data, oolo AI is an augmented analytics solution built with an understanding of how ads are used to generate growth and monetization. Critically, that foundation makes it possible for oolo to distinguish correlation from causation and to identify the real-world implications of a given data shift.
According to the company’s CEO, Yuval Brener, “oolo specializes in turning growth and monetization data into predictive and prescriptive insights”. It works by modeling and contextualizing your data and then plugging it into a predictive matrix. The output predictions work in parallel with an anomaly detection engine to surface the details of any occurrence that might open the door to fixes or optimizations.
Understanding the Demand
Subject to thin margins and sensitive growth models, publishers live and die by the whims of an ad landscape that is incredibly complex and prone to changes (e.g. ad network decentralization, the end of 3rd-party cookies, the discontinuation of IDFA tracking, the rise of ad blockers, etc.). To keep revenues flowing at full strength, millions of digital interactions (data points) need to be monitored through an increasingly complex and fast-paced ecosystem. The challenge is getting good information out of raw data and managing that information in a way that makes it easy to access and action.
It’s a challenge that’s compounded by the fact that most publishers rely on manual reviews — with the help of visualization tools — to hunt for insights. It’s a slow, tedious process. But with the speed of business being what it is, slow work tends to get stripped down and hollowed out. That results in monitoring that is even shallower and more error-prone.
Missed opportunities and overlooked problems quickly add up to stifle monetization and undermine growth. In the midst of a deepening economic recession, that’s a recipe for failure. But even when issues are caught, it doesn’t mean that all is well. The X-factor is the speed (or lack thereof) at which businesses are able to recognize and respond to those issues.
According to a market survey conducted by oolo, 64.7% of respondents cite manual data reviews as their primary means of detecting issues. At the same time, 53% say their biggest day-to-day pain comes from manual and inefficient processes. Those two facts taken together represent an obvious problem. Which is why workflow automation ranks as the top priority for growth and monetization professionals evaluating new business tools.
And it’s not just about the inconvenience or wasted time. It measurably slows business progress.
That’s the problem that oolo says it’s solving. It’s what they describe as “anomaly detection with a business brain.” Combining deep domain expertise with machine learning (predictive modeling and deviation analysis) and impact awareness, oolo covers both sides of the ad equation — the buy side and the sell side; serving marketers as well as publishers.
Rather than presenting a static snapshot of your business, oolo strives to create a living, breathing model that alerts managers to problems & opportunities and anticipates where performance is heading.
An Explosion of Possibilities
If you’re running an operation of any significant size and you’re relying on KPI dashboards, rule-based alerts, human data review, and manual investigations, there’s zero chance you’re not leaving value on the table.
While mediation and attribution reporting provide exhaustive data access, oolo promises end-to-end performance supervision. Data access is crucial for when users want to proactively launch investigations and hunt for information, a priori. End-to-end performance supervision, on the other hand, is what’s needed to make fast sense of emerging UA and monetization trends while eliminating the risks of delayed detection and missed issues.
On the buy side, oolo helps marketers get the most out of their ad campaigns by maximizing return on spend and ensuring that nothing undermines growth. In practice that means continuous data analysis — probing, for example, which channel delivers the highest value users (highest CVR, ROAS, etc.) for each dollar spent. It also means tracking channel and cohort changes to alert you know when the landscape shifts and action is needed, or new opportunities emerge.
For app marketers, oolo is said to streamline workflows and enhance data responsiveness. It’s a shot in the arm that’s invaluable for day-to-day tasks as well as strategic imperatives. When it comes to scaling UA, for example, the added speed and precision with which channel investments are managed makes it possible to grow installs while holding CPI (cost per install) steady and maintaining or improving cohort quality. According to oolo in fact, its clients enjoy ARPU (average revenue per user) uplift of almost 30%, on average.
On the sell side, oolo helps publishers get the most out of their ad inventory by maximizing asset utilization rates and total monetization while optimizing revenue splits. The idea is to automate data monitoring and alerting across all ad revenue and ad revenue-related aspects of the business — whether its metrics like fill rate, changes in buyer behavior, versioning issues, data discrepancies, technical problems, or A/B tests.
The cumulative impact of such improved oversight is significant — with managers moving more quickly and more effectively from analysis to decision to action. According to oolo, it makes enough of a difference to result in average revenue uplift of 5-14%.
But the impact potential doesn’t stop there. The business modeling that powers oolo’s alerting can also be used to forecast revenue and set quarterly goals, or more optimally price products/inventory.
Regardless of the specific application though the point is the same: to zero out the space between information and optimization — drawing a straight line to ROI. As a successful entrepreneur and renowned marketer Neil Patel put forth at a recent Web Summit, the future of business demands more dynamic, business-aware, and interpretative analytics.
“[The greatest impact will be felt with] AI analyzing your analytics in real-time and telling you where there’s waste right then and there instead of having humans analyze your spending [only periodically] — once a day, once a week, or once a month. So then it prompts a human and from there the human is [able to take action,] telling you ‘Yes, let’s stop this spend here and reallocate the money’ or ‘Spend more here because it’s working’.”
It’s an alluring vision of the future to be sure, and it seems that the future is already here.
Hi, I’m Oren, founder at BIGINTRO, a content strategy agency that helps B2B companies drive growth. We develop search, social, PR, and content marketing strategies tailored to business goals. I also have a dog named Milo.