written by
Nils Brosch

Failing-forward through Lost-Deal Analysis

Sales Management SaaS Sales Process 4 min read


If you build your sales playbook on characteristics of 'won deals' only, you are blind to the reasons why most deals simply don't close - a clear case of survivorship bias. Analyze your lost deals to learn from your mistakes and then have a repeatable process in place to run experiments to improve on the status-quo of your sales process.

Lean Startup by Eric Ries

If you work in a somewhat innovative field or a startup, you will inevitably have heard of 'Lean Startup' by Eric Ries. Essentially Ries is advocating quick continuous learning by analyzing small scale experiments. Whilst CEOs and Product managers embrace this topic to a large extent, I have rarely seen sales managers thinking that way.

Applications in sales

The living sales playbook

If you want to make any sort of experiment in your sales team, you will need to make sure that you have a reliable and consistent basis to start iterating upon. For processes, this means that you have an overall consistency in your team's approach towards sales. Hence, make sure you have a living playbook that is used by reps and trained/checked upon by sales managers. Sales team consistency is one of the highest predictors of overall sales success: a. it allows for effective management through changes of best-practice and b. halos a certain maturity of processes. So what do I mean by living sales playbook? A playbook that is iterated upon continuously.


Once you have a playbook and consistency in your sales conduct. Let your team define a backlog of experiments that you run according to the model that Kyle Lacy from Lessonly described in the Predictable Revenue podcast. 1. Decide what to test 2. Document & plan the experiment 3. Deploy 4. Analyze and Adjust.

Failing Forward

Patty McCord contributed an interesting article on Gong.io's blog about the importance of doubling down on what works, whilst cutting the 'dead weight' of process inefficiencies. The latter is something that strikes me as something that is practiced too little in most companies.

Patty McCord: Culture Isn’t Built, It’s Empowered!

How to decide on experiments?

The first step of the experimentation phase is crucial. It is tempting to try out new stuff all the time, whilst ignoring the ineffective process slack that builds up over time i.e. visiting companies face-to-face instead of video calling.

If you build your playbook on characteristics of 'won deals' only, you are blind to the reasons why most deals simply don't close.

Survivorship bias

A popular analogy is the following: In World War II, allied forces studied planes that survived being shot to locate appropriate armor reinforcement. By studying surviving planes, they failed to build stronger armor in actual vulnerable areas that led to shot-downs/crashes of lost planes - survivorship bias. The same concept applies in sales if we only study won deals, instead of lost ones.

An example:

Let's assume that you are applying a variation of the 'Continuous Rapid Improvement Model' (see above) in your organization and aim to increase conversion rates in the bottom funnel with a new proposal tool like i.e. HelloSign. As a result, you get a 5% better conversion from Demo-to-Close, resulting in a CR of now 35%. High fives all around? Well, yes - celebrate the improvement, but realize that the biggest source of information will remain untouched and it is the 65% of lost demos that didn't convert.

We tend to copy the best-practices of our most successful reps, however, these are the embodiments and consequences of won deals, more than likely the majority of their deals actually didn't close. You must find out why.

Practical implications:

Assing a high priority to 'Win/Loss Analysis'

Be honest, do the lost reasons in your CRM look roughly like this?

In the end, this list is either never touched or simply displayed without comments to investors. First of all, these reasons above do not appear to accept responsibility for the loss of the deal. Secondly, the reasons don't tell us anything about what we can do differently next time to improve our chances. Let your reps fill out a separate free text field to state what they could have done to mitigate the stated lost reason. Furthermore, take time out of your schedule or your 1on1s to discuss 'lost deals' and fill the experiment backlog with new ideas gained from lost-analysis.

The big picture - building an antifragile sales process

This approach to learning from mistakes and living a 'via negativa' are two components of what is called Antifragility. In short, something is 'antifragile' if it improves under pressure/stress/adverse conditions. I am obsessed with antifragility since 2014 as I wrote my master thesis about the topic and I want to apply this crucial concept to sales in a book:

If you are intrigued by the topic and want to stay up to date on other related posts or developments with the book, please subscribe to the mailing list below:

Sales SaaS