For almost a decade, I had the good fortune of working for former McKinsey consultants. They taught me two simple, yet valuable, data-driven frameworks. Below, I’ll explain how we applied these frameworks to optimize sales engagement at Salesloft.
The First Framework: Issue Tree
The first framework is called an issue tree. Here, you start with an end-goal. Then, work your way down by asking what mutually-exclusive (i.e., independent) and collectively-exhaustive (i.e., complete) strategic levers drive successful attainment of the goal. Since collectively-exhaustive can be, well, exhausting, I push myself to think of 2 to 5 ‘children’ for each ‘parent’ in the issue tree.
Enough theory on this first framework. Let’s look at optimizing outbound sales development representative (SDR) performance using an issue tree.
Sales Engagement Issue Tree
Just as in writing a story, the hardest part of building an issue tree is choosing where to begin. To ensure clarity, use the verb + goal format when naming nodes. One could start a level up at ‘Meet/exceed targets for meetings held’ or two levels up at ‘Meet/exceed targets for sales qualified opportunities’ or higher still. To limit complexity, I choose to focus on setting meetings. One should select the top-level parent based on the goal that is most important and within your control to optimize.
Achieving a target for meetings set requires two things – high activity and effectiveness.
High activity is mostly a matter of will, a factor that has its own inputs and outputs. Were I focused on the inputs, I would have fleshed out the issue tree to include facets of intrinsic and extrinsic motivation. While motivation is critical to success, I decided to focus on two measurable outputs: engaging new contacts and hitting touch volume targets.
In contrast, high effectiveness results mostly from skill. As illustrated in Figure 1, effectiveness can be decomposed as follows:
- Engaging quality accounts
- Running optimized cadences
- Exercising selling skills
At Salesloft, our sales development reps do not have a lot of influence on engaging quality accounts. Our marketing team defines our ideal customer profile with input from sales leadership. Then, our revenue operations team identifies and tiers the universe of possible accounts. Finally, our sales development managers assign accounts to reps to ensure they are always working the appropriate number of accounts.
Similarly, our sales development reps have limited influence on which cadences they run since those are optimized via A/B testing. Our sales development managers set up and run their own tests, occasionally collaborating with our data science team. That said, our sales development reps can and do engage prospects above and beyond the minimum required. We refer to the totality of engagement as a “derived cadence.”
Effectiveness then comes down to each rep’s direct engagement skills. That covers email personalization, phone engagement skills, social touches, and so on. Our research tells us that sales professionals should spend about 5 minutes per contact to personalize the first 20% of their email copy, inclusive of the subject line.
The Second Framework: Bright Spot Optimization
Now that we have the first framework in hand, we turn to the second framework, bright spot optimization. The idea here is based on the observation that there is natural performance variation in any group of people, even if they receive the same information and training.
The bright spot optimization recipe:
- Select a measure that is as objective as possible.
- Stack rank people on that measure.
- Identify the difference in behaviors between the high and low performers.
At Salesloft, our preferred measure of sales engagement effectiveness is the number of sales qualified opportunities (SQOs) generated per 1,000 activities.
In Figure 2, the y-axis represents the number of SQOs per 1,000 activities, and the x-axis represents the number of activities. Each dot is an individual salesperson.
The upward slope of the line means that reps ride a learning curve. Predictably, with more activities, they become more effective. The R-squared of 0.25 means that 25% of the variation in SQOs-per-1K activities is explained by a variation in the number of activities. In other words, a quarter of the variation in sales effectiveness comes from on-the-job experience. The other 75% of the variation in sales effectiveness comes from other factors that may be within or outside the control each rep.
Now, we move to step 2 of the recipe, stack ranking. There are at least two approaches. One is to stack rank without regard to how people should be performing given their tenure. If we were to do that, due to the learning curve, we’d mostly uncover the things the strongest tenured people do that the weakest newcomers do not. While that is not a bad approach, we add a slight twist.
People who are below the line are underperforming relative to where they should be. People who are above the line are over-performing relative to where they should be. We stack ranked based on distance above or below the line. That way, we look at high achievers independent of tenure.
On our team, we found a 10x delta between our highest and lowest performers, though most were clustered around the average. Statistically speaking, rep performance roughly follows a normal distribution.
It is important to note that we did not hold the under-performers at fault. It is just that the over-performers happened to figure out something special on their own. Our job as leaders is to discover those behaviors and teach them to the under-performers. Under-performers are only held accountable after leaders have done their job, training and coaching their people.
With our stack rank in hand, we next interviewed the top and bottom 10% to understand what they were doing differently. During interviews, it is especially important to ask people to describe what they actually do rather than what they would or should do. Aim to withhold the stack rank information from the interviewer so that they are as unbiased as possible. Otherwise, they may unconsciously home in on confirming information or unintentionally ask leading questions.
At Salesloft, we found three things our top performers did more consistently.
First, our most effective reps set up a ‘hit-list’ cadence that they run in parallel to our required team cadence. This is a one-step cadence that contains their highest priority contacts. Whenever an over-performing rep had extra time or resources to spread some #SalesLove, they run through the hit-list without marking the given step complete. Their activity is still automatically tracked in Salesforce, but the contact remains on the hit-list for extra engagement in the future. Observing this, we set up a team cadence so that all sales development reps could adhere to this best practice.
Candidly, we were conflicted when we found out about this hit-list cadence. Our first inclination was to simply bulk up our team cadences with the extra touches of various types the overperforming reps were doing. We decided against this course of action for three reasons.
One, these extra touches are suitable for our highest priority contacts only. Two, the number and types of extra touches vary by contact. And three, perhaps most importantly, the point of bright spot optimization is to replicate a best practice. If we modified the hit-list approach prior to implementing it, then we would never be able to judge success or failure. Moreover, we can always refine our approach later. Creating cadences for high-priority and regular-priority contacts, for example.
Second, our top performing reps were more consistent about leveraging referrals from senior co-workers. We discovered that, as our company has grown, our newer reps were not as comfortable asking for help. We quickly turned things around by reminding our reps that senior folks love to be leveraged. Though it has become more challenging as we have grown, we strive to give our reps as much direct interaction with senior leaders as possible.
Finally, our most successful sales development reps were much more likely to call prospects when they saw high client engagement with our prospecting emails. They monitored their activity feed and called quickly after seeing a spike in activity. We institutionalized this best practice in our Salesloft instance, using automation rules that trigger cadence steps in response to activity spikes.
I imagine these three best practices will immediately accelerate your success in outbound sales engagement. However, consider building out your own issue tree and conducting a bright spot analysis to figure out what your stars are doing. This way, you can elevate the performance of your entire team.
Interested in learning more about cadence best practices and benchmarks? Check out our latest report here.
