Understanding the ROI of Win/Loss

The ROI of a Win/Loss program depends heavily on the quality of the program you build. Build it well, and the ROI can be significant.
by: 
Brennon Garrett
Kaptify Founder
Brennon has conducted thousands (and thousands) of Win/Loss interviews. If he doesn't hold the world record for most Win/Loss interviews ever conducted, he's at least a contender.
Text Link

Setting the right ROI expectations

There's a popular Win/Loss  quote out there from Todd Berkowitz of Gartner Research that says “a good Win/Loss program can improve win rates by up to 50%, and revenue increase 15-30%”. Those are some pretty exciting numbers. But... are they true? Can you Win/Loss actually grow revenue and win-rates by that amount with Win/Loss? 

We run a lot of Win/Loss programs for our clients, so we have a front-row seat to the impact a good program can have on revenue and growth. Our opinion is that well, yes, Todd is basically right. But if you're going to have a chance at hitting those numbers, we need to add a lot of context. First of all, Todd clearly isn't saying it WILL grow revenue by that amount, he's saying that it CAN grow revenue by that amount. Those are very different words. The whole ROI question swings on his use of the word "good", and how we define "good" in this context. For what it's worth, we've seen a lot of Win/Loss programs out there in the wild, and I assure you, most of them are far from "good".  They're quite bad, actually. Why? Because Win/Loss programs are  deceivingly complex beasts. Very few teams understand what they're getting into when they decide to take one on. Building a good program takes  5-10x more time and energy than most teams estimate.

The amount of time and energy you put into building and running your program is almost perfectly correlated to the amount of new revenue it will generate.

Unfortunately most teams underestimate the time and energy it takes to build a good Win/Loss program by 5-10x.

Revenue growth from Win/Loss = Total Time & Energy Invested into program

If you're interested in knowing how much work it takes to build a good Win/Loss program, check out this article here. And don't worry, if you're under the impression that setting up a Win/Loss program shouldn't be that hard, you're not alone. It's a cognitive bias shared by almost everyone we've spoken to. One of the reasons people tend to underestimate what it takes is because Win/Loss programs have so many moving parts that most of them are below the surface and not visible until you're well into the process.  And then once you understand how much additional work is required, it becomes obvious you just don't have the time. It's de-motivating, and the program quietly slides into poorly managed and de-prioritized initiative. Incidentally, this is one of the reasons a 3rd party can be helpful.

For this article let's assume you've accepted the idea that building a Win/Loss program takes a ton of work. And that what you're really interested in is how much new revenue a well-run Win/Loss program can actually produce?  

How does Win/Loss actually increase win-rates

At first glance you probably have thoughts like "well, if we can figure out what we're doing wrong, we can simply fix those things and our win-rate will go up." And yes, that is theoretically true. But generating new revenue from a Win/Loss program is all about identifying what needs to be fixed, and then the real-world messiness of actually getting them fixed. And doing this at scale. It's an exercise in building a machine that consistently identifies what needs to be fixed, and building workflows to get as many of those issues fixed as resources allow.

Win/Loss data is very good at figuring out what needs to be fixed. But someone actually has to do the fixing. Someone has to have the conversations, to get buy-in, to allocate resources, and to drive outcomes.
Win/Loss is a volume game

There are no silver bullets when it comes to driving new revenue with Win/Loss. It's a grind, and it's a volume game. You need to surface a high volume of insights, and you need to fix a high number of those issues. It's very rare that you can jump in and find a single issue that will suddenly drive a huge amount of revenue. It can happen, but it's rare. "The Tortoise and the Hare" is an apt analogy for Win/Loss. Most people jump in and start acting like the Hare. But to drive real growth you need to be the tortoise.

There's an interesting phenomena that we've observed many times once people realize how much  new revenue a Win/Loss program can unlock, they tend to be unrealistic about how quickly they should be able to unlock that new revenue. Executives are especially vulnerable to this because they're always under a lot of pressure to grow revenue. Done well, Win/Loss well can indeed create a lot of new revenue, but it takes a lot of work, and it takes at least it takes 6-12 months to start seeing gains

There are no silver bullets in Win/Loss Analysis. Driving new revenue with requires identifying lots and lots of small issues, and fixing them one by one. Once you've fixed enough of them, you'll start to see small upticks.
It takes 6-12 months to start seeing gains

The reason it takes 6-12 months to start seeing gains is because you have to fix a lot of issues before you start seeing upticks. And to fix a lot of issues you have to build a complex machine that gathers all the data, derives insights, distributes them to your team, and then your team needs to action them. That's a lot! If there's anything I've learned from doing Win/Loss Analysis it's that identifying an insight is worthless if it's not actionable.

The actionability of the insight drives growth and ROI far more than the insight itself. The magic is mostly in the actionability.
You have to build a big machine

To build a successful Win/Loss program you have to build a big machine that does a lot of things and has a lot of parts. Some of the big parts of that machine include recruiting participants, conducting interviews, analyzing interviews, deriving insights, and presenting your findings to teammates. At first glance, those parts seem straightforward enough. But as you press into any of those areas you start to see a lot more moving parts. Let's take the "analyzing interviews" part of the machine and just crack it open a minute.

When you look at all your "reasons for loss" data, the reasons will usually fit into the following categories: 

  1. Product issues
  2. Pricing issues
  3. Sales team issues
  4. Competitor behavior
  5. Customer issues (things like budget cuts, personnel changes, etc)
Example: How the machine works to fix a sales team issue

Let's say you can tell there are some issues with how the sales team is selling the product. So you decide let's try to fix issue, and maybe win rates will got up. Sales issues tend to be the easiest to understand, and often times the easiest to fix. But they still take a lot of work to figure out and fix because sales issues usually occur at the individual sales-person level. The sales team is not one person with one set of behaviors. It's a group of unique individuals all with their own unique personalities and styles that are impacting sales in their own way. To successfully push through a Win/Loss fix through the sales team you need to achieve the following:

  1. Identify a pattern: A common patterns is that a single rep may be doing something that is turning off potential customers. In order to identify this pattern you need to conduct enough interviews to identify the pattern (if you have a sales team of 5 people, you probably need 25 interviews before you have at least 5-6 loss interviews for this sales rep)
  2. Present the data: That data needs to be gathered, organized, and presented to the sales rep's boss, and you need to get buy-in from the boss that there is indeed an issue and that they'll share it with the rep.
  3. Get buy-in from the sales rep: That sales person needs to see and understand the issue, be willing to change, and then to actually change.
  4. Repeat steps 1-3 (multiple sales reps) for scale: To see any measurable revenue growth, steps 1-3 need to occur at scale, and therefore need to occur with as many sales reps as possible. It's easy to identify issues with one rep. But you'll probably need to repeat this process for at least a handful of sales reps before you start seeing upticks. And to get to that point you need to conduct a lot of interviews.
Example: How the machine works to fix a product issue

Fixing product issues with Win/Loss data is a totally different animal. First of all, product issues tend to dominate the other categories when it comes to "why we're losing deals". And secondly, products are complicated. So the product issues that surface during Win/Loss interviews can also be complicated. Some of them will be quick and easy fixes. Some of them will be complicated and take 12 months of developer time.  Some insights will drive value for certain customers while other insights will drive value other customers. There are lots of interesting tradeoffs when it comes to making product decisions with Win/Loss data.

Driving revenue by fixing product issues requires you to build a product-decisioning-schema to understand all the tradeoffs. With a good enough schema you can start making very smart recommendations to the product team.

Product schema considerations:

  1. Customer segmentation. Which customers segment is this for? Does it benefit certain customers but not others? And if so, does that align with our revenue goals?
  2. Measurability: Does the potential revenue impact of this feature have high or low measurability?
  3. Developer Resources: How much dev time will the build require?
  4. Data confidence: How much Win/Loss data do we have to support our conclusion?
  5. Product roadmap alignment: Is fixing this issue reasonable aligned with our product roadmap and high-level product goals?
  6. Product team alignment: How well does this insight align with what the product team already believes to be true? (i.e. how much selling will you need to do?).

An additional wrinkle in this picture is that if your product is a piece of software, gathering good product data is hard. Software products have very large surface areas, and interviewees talk about lots and lots of different parts of that surface area. Usually Win/Loss feedback is helpful on the best and worst parts of your product. But for everything else in-between you'll need to let your product analytics lead the way.

High revenue growth but low measurability.

If you put enough work into building a robust and ongoing Win/Loss program, you're going to drive a lot of new revenue for your company. But don't expect precise measurability. Is it better when things are measurable? Yes. Can you measure how much revenue growth your Win/Loss is driving? Uh, not so much.  

As you can see from the examples above, it's very difficult (if not impossible) to measure the revenue impact of each little insight. And at the same time it's totally obvious that if you gather enough good insights and fix them it will dramatically grow revenue over time.

Win/Loss insights usually have a "common sense" kind quality to them. You can look at them and immediately see that actioning them will drive new revenue. You can argue about the details of which insight is likely to drive more revenue. But your common sense is usually a good enough guide to figure out how to prioritize the insights.

If you want to track revenue growth and win-rate increases, our advice is to track how many insights you've identified and actioned over a 12 month period, take a look at revenue growth over that same period and subtract the obvious new revenue drivers at your company (new sales hires, launch of new products etc). From there, do your best to approximate the impact of Win/Loss insights on revenue growth, and be clear that it will be an approximation.
High revenue growth and high measurability.

Interestingly, there are two very different "types" of insights you get from Win/Loss interviews. And they're totally different from each other when it comes to measuring their impact on revenue. One is not really measurable, and one is totally measurable.

  1. Company Level Insights. These are the large trends and patterns in your data. If a bunch of people say your price is too high, you have an aggregate insight. These insights are structural in nature and tend to point to a large issue in your pricing model, your product, or your sales team.
  2. Customer Level Insights: A customer level insight is an insight about an individual account (quick note: most of our clients are B2B SaaS companies with an ACV of at least $10k per year). If on average, an individual account drives enough revenue for you that it justifies time and attention from your team, then a Win/Loss program can generate a tremendous amount of these types of insights.
    1. Winback Customers. About 25% of the "loss" interviews you interview have the potential to re-enter the sales pipeline because of your Win/Loss interview. If you that customer is worth $15k annually, and they re-enter the funnel and then get closed because of your Win/Loss conversation, that event is super measurable. And if you conduct enough Win/Loss interviews over the year, that number will start growing quickly.
    2. At-risk Customers. Amongst our clients about 20-30% of the "win" customers we interview have significant risk for churn, and that risk gets revealed during the Win/Loss interview. If you can uncover what the customer's pain points are, and funnel those issues back to the account managers in time to save the account, you're effectively saving this account from churn and capturing revenue that would have been lost. This is a very measurable event.
    3. Upsell/Cross-sell customers. A Win/Loss conversation can be structured to do a little up-selling and cross-selling. It takes a little work on the front-end (i.e. you need to think through how to ask the questions without sounding too "salesy"), but you can use the Win/Loss interview to start generating upsell and cross-sell leads for the sales team. On average about 20% of win interviews tend to have upsell/cross-sell potential. And once again, this type of event is very measurable.
If you want to justify your Win/Loss program to executives, or to the stakeholders on your team, show them ROI data from Winback, At-risk and Upsell customers. It only takes closing a few of these accounts to justify the cost of the the program for an entire your.

If you build it right, the revenue will follow

In conclusion, Win/Loss programs can clearly drive a lot of new revenue for teams that spend the time and money to build them well. But instead of getting too caught up in the details of exactly how much growth Win/Loss can drive, think about the essence of what Win/Loss insights actually are (asking people why they didn't buy your product). It's super valuable data, and if you do it well, it's going to drive a lot of revenue. Instead, think about how to build a good program, and how to put in motion the things you'll need to build it successfully. The documentary filmmaker Werner Herzog has a great quote about raising money for a film, which is analogous to driving growth with Win/Loss data:

“If your project has real substance, ultimately the money will follow you like a common cur in the street with its tail between its legs.” --Werner Herzog

If you build a Win/Loss program that has real substance, the growth and revenue will follow.