Reverse Funnel Analytics

What happens when we fall short of our sales target? Do we try to understand why, and why we didn’t know earlier? It’s difficult, especially in these dynamic times, to predict what will happen. With reverse funnel analytics (RFA), it is possible to remove a great deal of the guesswork.

A growing number of organisations take a data-informed, lead scoring approach to resource allocation, scientifically calculating how many, and what type of leads they require to convert into opportunities and ultimately into sales to hit their targets. They know, in detail by month, whether they are on track thanks to these solidly calculated early performance signals, often known as leading indicators. They use these insights to meet targets as effectively and efficiently as possible. However, such data-informed approaches often uncover unpleasant surprises, such as poor early funnel health, or historical insufficient velocity. They may even realise that targets are too unrealistic!  We address this challenge with our reverse funnel analytics approach, where we use mainstream lead scoring on a whole new level, to help organisations see exactly the volume, value, and type of leads they will need – and crucially by when – to hit their sales target.

Lead scoring is a baseline methodology for resource and performance optimisation

Scoring your leads with machine learning algorithms is a way of working that many organisations currently consider or use. Companies face the challenge of reaching high sales targets at the minimum possible cost and resources, meaning one simply must prioritise the leads with the highest likely conversion rate. Marketing and sales teams can use the lead scoring model’s outcomes to prioritise and plan their customer engagement.

Reverse funnel analytics takes lead scoring to the next level, an opportunity that many organisations miss

Some, probably most, organisations have a ‘rule-of-thumb approach’ towards goal setting. Teams estimate what they can achieve based on business judgement and experience, sometimes using data to guide this decision making. Especially now, unexpected external factors as well as internal changes mean that these goals are not always set up for success.  Reaching end-of-year performance, is a long journey with many milestones and hiccups down the road. How often do we install early warning signals to informs us when deviations from the plan occur? Reverse funnel analytics allows you to determine such early warning signals – often known as leading indicators – to predict, measure and course correct while you still have time to change the end result.

Reverse engineering process is straightforward in principle but hard in practice. For example, if a company wants to attain 20K followers as a result of a digital marketing campaign,  then 100 campaigns must be planned and executed this year because last year the company attracted 10K followers after launching 50 campaigns. But is it that simple? We are here assuming all conditions remain the same. In the context of sales performance, as with any other mathematical problem, it can be done using simple calculations, relying only on historical performance, or using data science models to create a predictive outlook and evaluate the reversed funnel, providing advanced insight.

The most accurate way to execute this is by relying on the scores from your lead scoring model. Reversing the funnel in a data-science way means utilising better lead scoring models, where apart from simply prioritising leads’ nurturing daily, it enables understanding of what is required every step of the way for teams to reach their yearly sales goals. The insights are based on the funnel’s health, meaning the quantity and predicted quality of the leads in the funnel, at any point in time. Factors like business seasonality, historical events, size of the team, etc. can all be integrated into the modeled expectations.

The reversed approach answers questions like:

• How many opportunities, with what probability of conversion, are needed today/this week/this month, to hit the order target?
• How many leads to hit the opportunities target? How many campaigns to get to a certain number of leads?

Consider the following situation: It is the month of July, and so far, Marketing and Sales are doing well, in terms of generating leads, moving them through the funnel, and closing deals. However, based on predictions, the performance is expected to drop in September, generating a significant gap between expected and actual results. The immediate question is, how much better should the team be performing, to avoid this drop? The predicted reverse funnel would show that the volume of leads entering the funnel is ideal, but the MQLs and SQLs are staggering behind by almost 50% (check the visual below). This means that the leads are of good quality, but they are not moving fast enough through the funnel and, without reacting on time, they might convert in September with a lower-than-anticipated performance. The immediate action is to shift more efforts to nurturing the already existing leads.

RFA is also about checking If your targets are realistic. Essentially can you achieve the targets with the leads currently in the funnel and the velocity that they are moving with? Having a forecast that predicts where you will end up if you continue to perform the same way, based on previous data which considers historical trends, seasonality, and the progress in the current year so far, can you reach targets, or do you need to adjust them.

Imagine, you have been given targets by the higher management and based on experience, the team sees them as being too unrealistic. RFA can help you, evaluate without a bias, whether this is the case and by how much the targets deviate from what is realistically possible. The conclusion will not only be based on what happened last year and the years before that, but also reflect on the size and performance of the team, seasonality, predicted conversion of the currently existing leads, etc. For example, if the target this year is $25mln, it might turn out that $18mln is what can be achieved in reality because of how the current and predicted funnel size and quality have been and are progressing over time. This conclusion can result in adjusting the targets or changing strategy down the line like increasing the sales or marketing team size, provoking additional analysis on what works best for attracting new leads, increasing advertising spend, etc.

Bottom line is, RFA provides a transparent and realistic outlook of your funnel, calculating the in and out of how the organization is expected to perform and most importantly, allowing for a timely reaction.

Embedding RFA into the organization requires change in mindset and technical execution with high accuracy

The initial challenge that might arise when implementing reverse funnel analytics would be establishing the funnel itself because it varies depending on the industry. A deep understanding of the company’s needs and business environment must be evident. For instance, if a marketing team fails to run an X number of campaigns, what other channels and ways exist that could still generate leads.

The traditional way of measuring goals is top-down and the real challenge of introducing RFA comes from its bottom-up approach. And this change of approach translates into a change of mindset, looking at the overall performance in a reverse way and keeping in mind that the whole process is unidimensional.  Meaning that if a particular KPI is not performing, then the next KPI in the chain will stagger in performance because they are all interconnected in one way or another. Understanding this “network of KPIs” is a crucial necessity and it is something still under exploration by today’s business environment.

Additionally, building the specifics of the context into the lead scoring model and the RFA is vital for the accuracy of the predictions. Setting up the change is a collaborative exercise that needs to involve business context from all levels, together with strong data science expertise, to ensure the extraction of maximum data value.

Once the models are in place, maintaining its performance is key but also building the new habit of constantly comparing ‘where we are’ vs ‘where we should be’, across all levels of the organisation.

Conclusion

Reverse funnel analytics is an out-of-the-box usage of lead scoring, reverse engineering targets and helping teams keep track of whether the sales funnel is moving as needed or adaptations need to be implemented.

This application of a fairly mainstream solution is different and appealing in that it enables reflection and learning as well as forward-looking prediction: in doing so, it connects the chain, starting with the lagging target, usually sales, and tracking all the way back to the leading indicators that will make the difference to hitting or missing targets. Reverse funnel analytics provides an executable manoeuvre for each person in the chain.  It’s a strategic optimisation tool that, when used well, enables learning, helps teams to prepare and predict performance, and ultimately safeguards a company’s bottom line.