What if we told you your business decisions are relying on 1-3% of your users? You probably knew that already, but were you aware of the dangerous bias of this outlook?Those 1%-3% are not randomly sampled users, but users that showed certain actions. How confident do you really feel about the outcomes of your strategy and tactics?
That’s the exact problem with survivorship bias.Survivorship bias is a cognitive bias that occurs when we focus on the success stories of a particular group, while ignoring the failures.During World War II, the USA Air force wanted to decrease the rate of fallen aircrafts. To do so, they ran statistical research on the aircrafts that returned from battle - examining the damages done, add armor to the areas that were hit most, and by that increase their durability in battle.A map of hits to the aircrafts:
Coming up with great ambitions, they fortified those areas and went into the next challenges with a very optimistic outlook.
The result: Planes kept crashing at a similar pace.
When the Statistical Research Group at Columbia University examined the damages done to the aircrafts, they came up with a bright idea:************Maybe we are looking at the wrong samples of planes, the planes that returned are not the issue - they did not take any critical damage, in contrast to the ones that crashed.
Their proposition - armor the planes in areas that were not damaged in returning planes.
The result: A dramatic decrease in the number of fallen aircraft.
Today, measurement and attribution tools rely heavily on conversion events to understand which users are the “best bets” - the ones you should look for in order to bring profitable growth to your organization. A best case scenario would be around 3% of your users. That means that because 97% of users did not make it to the finish line - you completely ignore them in your decision making, and you minimize your available audience pool, missing out on a LOT of potential buyers, revenue and business opportunities.
As we know, the user shopping journeys are complex and diverse, and could be impacted by variety of variables: Great potential customers could’ve dropped because of technical issues between the device and the website, others could tackle missing inventory and sizes, or maybe they were just 1 category page away from finishing their purchase. Ignoring the users that did not convert, and looking away from non-complete journeys would lead the User Acquisition optimization algorithms to focus on a narrow group, with limited potential and even lower profitability.
If you haven’t started collecting user behavior data on your website, you need to implement a first-party data collection solution ASAP.
Today’s solution look at the journey from a very simplistic view - focusing on the pages you visited and items viewed. While providing insufficient information to make high accuracy and complex decision, this would be a great first step.
Analyze your data, look into different channels and what types of users they bring to your website. See what behavioral patterns different users express, find out which journeys are successful - try to really understand what made them stand out from the rest and what happened to users with similar journeys that did not convert.
Instead of relying on data from single digit percent of users, start looking at more users and more website actions.
If you would like to take your solution to the next level - you would need to start looking at deep behavioral data. While item views and add-to-carts provide information, they are just scratching the surface.
Deep behavioral data is the most granular way of looking at user behavior on your website. We understood that since the user journeys are getting shorter and shorter, and as privacy restrictions are prohibiting data sharing and retention, the ability to understand the user behavior using the existing tools is just lagging behind.
By inspecting every micro-interaction of the user with the website, we generate X15 more data points, and using our cutting edge novel models, we can provide much more information on every user, starting from the very first moment the start their shopping journey.
Using Kahoona’s deep behavioral data, we can now get a much better understanding of our users, which channels and campaigns bring us the best users and the best business driving results. Since the data is available in real-time, it can supercharge your personalization engines using high quality data, with incredible scale.
Shift your spend from irrelevant users to high-value users, reduce CAC and improve your conversion rates.
Interested in understanding your users better, cut-off wasted marketing budgets of irrelevant users, and draw in much more of your high value audiences?
Schedule a demo with us