If an anomaly occurs in a KPI but your analysis tools are unableto spot it, did it really happen?

Today almost every business is a data-driven business. Whetherit's the monthly scrap rate at a factory or the conversion rate ofwebsite visitors to sales, each business has one or more criticalmetrics. These key performance indicators (KPIs) are indicators ofthe success and health of the company. Each data point from a KPIis a snapshot of that particular quantity at that particular time:The factory scrap rate for March, or last Tuesday's percentage ofwebsite visitors who actually bought something.

Once You've Got It, Plot It

Since each data point is a snapshot, one of the first thingsorganizations do with the KPI data they collect is plot it overtime. The resulting time series can answer at least a couple ofquestions, such as, “Are there any detectable patterns?” and “Isthis metric going in the right direction?”

Often the answer to first question (even if it's a “yes”) is notvery useful from a business standpoint. That's because even if youcan detect, for example, a day/night pattern in a businessmetric, there's usually not much that you can do to change thatpattern – the length of a day is stubbornly resistant to humaninfluence. Since the only metrics you can improve are the onesyou're able to change, detecting a pattern like this is useless forproviding actionable insights.

Unless of course you're able to relocate to Mars, in which caseenjoy your extra 40 minutes.

The answer to the second question is found by looking at thetrend line, the general slope of the plot over a long period oftime. Trend lines are important and informative, but they also havetheir limits. Perhaps the biggest limit is the fact that they areinherently long-term indications of data over time. By design,trend lines smooth out variations in the data, like a movingaverage.

A Trend Line Is Not the End of the Line

When those variations are the normal variability in the data,that smoothing is welcome. When those variations are a bend in thecurve, the start of a new “normal,” that smoothing will hide animportant turning point. Used indiscriminately, trend lines canactually prevent you from extracting insights from your data, andas you can imagine, hidden insights lead to missed opportunitiesand lost money.

That lost money can take the form of missed revenue because youhad insufficient inventory of a product that suddenly surged inpopularity. Another example is an uptick in users leaving yourplatform due to a brand-devastating “babby”-grade meme gaining circulation on social media.

Both of those examples require quick action, but you can't acton something you don't know is occurring, and you don't knowsomething like this is occurring unless you can detect it in thedata. Trend lines are useless for detecting business incidents likethese because it can take a lot of unusual data points toaccumulate before the trend line significantly changes.

Meanwhile, you're hemorrhaging revenue and users. To stop thebleeding, you need a rapid response, and that requires quickdetection of these deviations in the data (also called anomalies)long before they shift the overall trend. This is why savvycompanies are moving to real-time anomaly detection.

Recommended Upgrade: Anomaly Detection

But what is anomaly detection? From our experienceat Anodot, a real time business analytics company and vendor ofanomaly detection systems, anomaly detection accomplishes threemain things:

• Creates a model of “normal” values for a given business metric,including any patterns that may be present in the data;
• Uses the model to predict what the next data point should be;and
• Determines if that next data point matches what the modelpredicted, or is an anomaly.

This three-step process naturally lends itself to automation viasoftware, which is a good thing because in some organizations, thenumber of monitored business metrics can number in the thousands oreven millions. Another reason why anomaly detection wins over trendlines: An anomaly doesn't have to be drastic in order to becostly. Detecting subtle anomalies requires sophisticatedAI-powered software that catches these events, yet doesn't floodyou with false-positives.

Your data can only reveal as much as your analysis tools allow.Basic tools like trend lines can mask significant events until it'stoo late. Anomaly detection can produce much faster results, and doso at the speed of today's business environment.

Rebecca Herson is Vice President, Marketing for Anodot.She can be reached at [email protected].