Almost all the metrics we use today have one common thread. Almost all of them look back to the past.
If you want to deepen the impact of your data within your organization, and expand your personal impact, you should focus 30% of your analytics efforts on using positive metrics.
Predictive indicator!
But first, let’s take a small step.What are metrics
Here’s the metric definition from my first book:
Simple enough.
conversion speed. number of users. bounce rate. All indicators.
[Note: Bounce Rate has been banished from Google Analytics 4 and replaced with a compound metric called Engaged Sessions – the number of sessions that lasted 10 seconds or longer, or had 1 or more conversion events or 2 or more page views.]
The three metrics above are historical. They tell us what happened in the past. This is true of almost everything (if not all) that you are reporting.
But who doesn’t want to see the future?
yes. I can see your hands
The problem is that it is difficult to predict the future. What is the quote…No one went bankrupt by predicting the past. 🙂
Why use predictive metrics
As analysts, we transform data into insights every day. Amazing. Only some of these insights turn into action for a variety of reasons (your influence, the quality of your insights, an imperfect story, etc.). Sad face.
One of the most effective ways to reliably translate insights into impactful business action is to predict the future.
Consider the following insights from the data.
Email campaigns have a conversion rate of 4.5%, double that of Google search.
Consider this:
our analysis suggests that you can go from 6 email campaigns to 9 email campaigns per year.
Finally consider this:
〇Based on your analysis, you know that you can go from 6 email campaigns per year to 9 email campaigns per year.
We expect this to generate an additional $3 million in incremental revenue.
The predictive indicator is New Incremental Earnings. Not only that, we used sophisticated mathematics to identify how much of the projected revenue would be incremental.
In which of the following three scenarios will Insights execute reliably?
yes. Those with predictive indicators.
Because it’s hard, really hard to ignore your advice when you’re making $3 million in revenue!
Start the predictive metrics journey: Easy peachy lemon squeegee.
In a delightful development, every analytics tool is adding predictive metrics to its arsenal. As a way to differentiate your company with a unique take on this capability, and as a way to bring tremendous value to businesses of all types/sizes.
In Google Analytics, the early predicted metrics were: conversion probability.
Simply put, conversion probabilities determine Likelihood of users converting Between next 30 days!
I was so excited when it first came out.
In this case, Google Analytics analyzes all first-party data to identify behavioral patterns that lead to conversions. Then it looks at all the people who didn’t convert and scores them on your behalf from 0 (no chance of conversion) to 100. (Very likely conversion).
Phew! That’s a lot of work. 🙂
Of particular interest is that conversion probabilities are calculated for individual users.
GA: You can easily access the report under Audience > Behavior > Conversion Probability.
An obvious use for this predicted behavior is to target remarketing campaigns to people (7,233 in the above case) who may need to be nudged to convert.
However, there are ways to further use this data to determine campaign effectiveness.
For example, here are the sources of traffic sorted by: average conversion probability…
In addition to understanding your conversion rate (last column), you can also consider the number of users arriving through that channel that are likely to convert in the next 30 days.
Perhaps even better, you can use this for segmentation. Example: Create a segment with Conversion Probability > 50% and apply it to your Favorites report, such as the Content report.
There is much more to explore.
[TMAI Premium subscribers, to ensure you are knocking it out of the park, be sure to review the A, B, O clusters of actionable recommendations in #238: The OG of Analytics – Segmentation! If you can’t find it, just emial me.]
Bonus Tip: I can’t recommend you get enough Accessing the Google Merchandise Store A Google Analytics account. This is real GA data that works well in real business and is well implemented. Access is free. Very good for learning. The screenshot above is from that account.
Threee Awesome New Predictive Metrics!
Everything turns upside down for an exciting world Google Analytics 4 It gives you a little more to add to your arsenal of predictive metrics.
conversion probability is Discontinued I’m using the GA 4, but don’t worry as you get similar type replacements. Purchase probability
The purchase/ecommerce_purchase and in_app_purchase events are currently supported.
You can do all the same things that we discussed above for conversion probabilities.
To get closer to your finance team, you need to be their best friend! – And get predictive metrics they’ll love. Revenue forecast
Expected revenue from all purchase conversions in the next 28 days from users who were active in the last 28 days.
Let your imagination run wild with what you can do with this power.
I encourage you to look at this forecast and brainstorm with your marketing team on how to overcome your revenue shortfall. Not only do we use paid strategies, but we also use earned and owned.
On the rare occasion that your revenue projections exceed your goals, you can use your vacation time to visit Cancun. (Wait. Skip Cancun. That brand is tainted. 🙂
There is another predictive metric that I always look forward to. Churn probability.
The probability that a user who was active on your app or site in the last 7 days will be inactive in the next 7 days.
what is that quote? Does it cost him 5000 times more to acquire a new user than to keep an existing one?I may be exaggerating a bit.
Especially for mobile app/game developers (or content sites, or entities where recency/frequency is a death-or-death proposition). Churn is a constant obsession and now you can get churn probabilities aggressively. Make understanding behaviors, sources, and users central to your analytics strategy.
GA 4 doesn’t just pass these metrics along. Algorithms need a certain number of users, conversions, etc. to do the right calculations on your behalf.
These three predictive metrics demonstrate the power of forward-looking calculations. To enable the company not only to look back (70% of the time you’re stuck), but also to look into the future (aim to spend 30% of the time), use these approaches. There is no limit to how far you can go. time here).
And consider segmenting Probability of Purchase, Probability of Earning, and Probability of Churn!
Bonus Tip: If you want to move to the free version of Google Analytics 4 and take advantage of the great predictive metrics above, here is a helpful article.
Predictive Metric Nirvana – Example.
For marketing analysts, there is little in terms of forward-looking forecasts with sophisticated analytics to help set overall budgets for the year. include Allocate that budget across channels based on the declining revenue curve and future opportunities, and forecast: sale, Unit selling priceWhen brand lift.
Here’s what it looks like from our team’s analytics practice…
Obviously all these cells contain numbers. You will understand that sharing them with you would be a career limiting move on my part. 🙂
We can say that there are 13 different sets of factors in this analysis (product launch, competitor behavior, historical analysis of effectiveness and efficiency, underlying marketing media plans, upcoming industry changes, and lots, lots, lots of data).
super cool – aka Carbide – Factors include being able to tie brand marketing to short, medium and long term sales.
The forward-looking allocation is based on simulations that can answer low, medium and high risk plans considering all of the above. Senior leaders can choose from which they believe they align with their strategic vision.
[Note: Strictly speaking what we are doing above is closer to Predictive Modeling, even though we have a bunch of Predictive Metrics. Potato – Potahto.]
We share our work as a way to solicit feedback on what we can do better. I hope it might serve as a north star if you start practicing predictive indicators.
In my experience, if you’ve ever felt like you as an analyst had no influence, if you felt that your organization was ignoring data, then you should look for predictive metrics to deepen your business influence and impact. There is nothing like
When people use faith to determine future strategy, one thing they miss is what impact faith-based strategies have. His last three lines above show how you stand out.
boom!
The danger of predicting the future.
you would be wrong
A lot at first. Then it gets better and better, and less and less over time as you anticipate the future.
(Machine learning helps us there because it can take in far more complex things and spew out scenarios we can’t imagine.)
But you are never exact. The world is complicated.
This doesn’t scare me for two reasons.
2. who is more right than you? Most modern businesses are run by faith. They typically use large amounts of data. It’s usually much better than faith. And if you’re wrong, you can actually go back and update your model (faith is usually not open to upgrades).
yes. Don’t be scared.
Every time you’re wrong, it’s an opportunity to learn and get better in the future – even if perfection is always out of reach.
bottom line.
My hypothesis is that I haven’t spent a lot of time on predictive metrics and predictive modeling. change this.
It’s a great way to contribute materially to your company. This is a great way to invest in personal learning and growth. A great way to ensure your career is future-proofed.
As an analyst/marketer, live in the future, at least for a while.
See you there. 🙂
As always, now it’s your turn.
Share your criticisms, reflections, tips, and lessons learned from the project.
