Win-Rate Analytics: Making Winning Repeatable
As an interesting experiment, over the last three weeks we spoke to over two dozen CROs, Sales Executives, Rev Ops leaders, and Financial Planning and Analysis leaders to talk about pipeline analytics. Interestingly, not a single one could tell us the probability of winning a deal by stage (never mind by team, sales rep, industry, product, time period, deal size, etc. - nobody knew the absolute number). We've all been guessing, but it's not our fault.
Example dashboard for fun here.
So, what happens when you don't know your true win rates?
- We apply our "Gut" instinct - which is really our pattern recognition of historical experiences
- We assume the future looks like the past - and without recent and constantly updating information, the timeframe we're using is often quarters to decades
- We coach our team members based on what we think is the right idea
- We make safe decisions because we don't have the data to back up the riskier ones - such as doubling down on a product, geo, strategy, or tactic
- We make large decisions slowly, based on macro-patterns and obvious movements, instead of quick micro-decisions based signals
As a result, operational costs are elevated while top line becomes harder to grow. Here's a few examples where guessing hurts the most:
- Coaching becomes less effective as it is often generic and lacking actionable stats
- Learning cycles are longer, and reps are left to "figure it out"
- The largest deals or the loudest reps get recognized, while quiet and repeatable plays go unnoticed
- Accurate forecasting becomes challenging, resulting in avoidable budget cuts and employee attrition
- Bottlenecks and sales process gaps remain hidden, leading to frustration and lost deals
- Product feedback loops are drawn out, and product-market fit is guess and check
Let's look at this fictitious heatmap of win rates by sales rep and by sales stage.
Starting from the top and working our way down, we can glean a few insights immediately that will help us coach reps and drive predictable revenue.
- Angelina: Winning 72% of deals in Proposal stage, which is low for our ideal customer profile
- Johnny: 71% Evaluation win rate and 57% Proposal stage win rate, deals won seem to skip Proposal stage
- Rihanna: Most consistent, need to work on disqualifying deals earlier and increasing win rate in later stages
- Shakira: Not putting deals in Salesforce, not progressing unless confident they will win
- Overall: Late-stage win rate is low, need to segment our audience to find areas of strength and weakness
If knowing your win rate is so important, why are we all guessing?
Most companies don't have the data science teams and tools to do the analysis correctly. These are the most common reasons we found:
- Lack of Data Science capabilities: True time-series analysis is required. Specifically, we connect into the audit log in Salesforce where even a small organization has hundreds of thousands of rows of changes. CRMs, ERPs, and spreadsheets are incapable of performing these calculations. BI tools are incapable of performing these calculations without a very skilled data-science operator.
- Calculation and Logic Flaws: Data science operators may have the tools and the skillset, but they don't have the deep domain expertise required to understand exactly how to calculate it. The most common problem we see is the most dangerous one - the data is presented beautifully, but the underlying analysis is flawed. See our DDM 2 blog for more information on this.
- Illusion of understanding: Leaders think they know, or at least they think that others know. Because they have access to what they think is good data and accurate insights, they stop looking and start decisioning. This can be really bad. See our DDDM 1 blog for more information on this. But while this is often true at the executive level, it's not true at all when you dig into the details. There hasn't been a single conversation without an "oh shit" or "eureka" moment (often both) when we show leaders insights from their data.
- Bad Data: Leaders think that they can't draw conclusions because the data is bad (Salesforce hygiene anyone?). While this is often true for most static reports, it isn't when you apply the appropriate data-science techniques. You can still learn immensely from what you have, as long as you understand the flaws in the data. For example, you can't find out where you are losing if lost deals aren't entered into the CRM, but won deals are almost always entered into the system for compensation reasons. So you can focus on where you are winning, and deduce where you are losing (based on where you aren't winning).
- Wrong Data Sets: Most people are using static reports (such as a Salesforce opportunity report), or at best snapshot reports, that neglect the important changes. For example, a Closed Won report pulled from the CRM has all the deals that were won over a given time period, but it doesn't contain any information about the path it took to become closed won - such the stages it touched since creation. Even a snapshot is missing the data between the two snapshot periods (often monthly snapshots are taken, and looked at on a quarterly basis).
What you can do about it:
- Utilize built for purpose analytical tools, such as our prepackaged win-rate analysis
- Hire an advanced Revenue Operations analyst with SQL, Python, or other developer capabilities
- Upskill your current CRM Admin or Sales Ops team
- Ensure your Data Science team has a sales subject matter expert
In Conclusion:
Knowing your win rate is critical to the success of any sales organization. When organizations don't know their true win rate, they rely on guessing, which can hurt both the top and bottom lines. Accurate win-rate analytics can make winning repeatable through improved coaching, faster learning cycles, increased forecasting accuracy, sales process optimization, and product alignment.
