This is the first of two posts that will examine whether teleprospecting is too complex for testing. I was compelled to develop this series after observing interaction on our LinkedIn B2B Lead Roundtable (which has rapidly grown past 8,000 members).
Members ask fascinating questions about teleprospecting that evoke thought-provoking responses. In fact, some questions will spark a furious debate that might go on for weeks or even months:
- Should teleprospecting reps leave voice mails?
- How many dials should someone make per day?
- What is the right level of compensation for a rep?
- How much of that compensation should be variable?
- What are the pros and cons of pay for performance from outsourced teleprospecting vendors?
- Is it better to outsource teleprospecting or bring it in-house?
- Should teleprospecting be a sales or marketing function?
As you can see, the questions can range from tactical to highly strategic. If the question is a topic I don’t know anything about, I can find myself changing my mind as I read one good argument versus another, back and forth, like a tennis match.
Some questions are not at all new. Like the voicemail question, the outsourcing question, the pay-for-performance question, or the dials-per-day question. People have been having these debates long before LinkedIn existed.
Many times, the more sage members will answer with caveats. “It depends” is a phrase that gets used a lot. For example, the optimal amount of dials per day per rep will depend on the solution’s complexity, the size of the potential opportunity, the rep’s familiarity with the accounts he is calling, and so on.
What is the best answer to a particular teleprospecting question?
Still, doesn’t it seem like there is probably an optimal answer for each situation for your business? Couldn’t we set up experiments to find out what works best instead of just debating the question endlessly?
Also, since there are so many questions, what’s best to test first? For complex questions, how do you structure tests that yield valid answers? After all, for B2B lead generation, the quantities are often minimal, and the buying cycle is very long.
The complexity of teleprospecting compared with other marketing functions
With email, you can create two different subject lines, randomly split your list, and figure out which subject line worked best. Email, landing pages, and direct mail really lend themselves to this kind of experimentation. The universe is often large enough to yield statistically valid outcomes. Important questions can be isolated and measured.
Teleprospecting could undergo this kind of experimentation. But experiments that use the scientific process require isolating variables that you intend to test. Everything must be identical between the treatment and control except the question you’re trying to answer.
So what if that question is “Which script will perform best?” Can you really isolate every other possible variable that could cause a performance difference?
What if one of your teleprospecting reps is good, two others mediocre, and a third pretty bad? And how much can you really control what reps say and ask? With teleprospecting, because people (that is, the reps) are basically the medium, testing is not easy to do well.
And it’s more than just the people. There are different compensation levels and models for reps, managers, and support functions. Each function has different ratios, e.g., the ratio of reps to field-sales professionals or the ratio of reps to the coaches that manage them. There are lower-cost and higher-cost labor markets.
There are also many approaches to onboarding, training, mentoring, and coaching reps. There are dials, connections, emails, and calendar invites. Then there are the numerous outcomes of these activities within the various stages of funnel progression.
The conversations are also complex. What do you say? To whom? When? There are questions and more questions for reps to ask. There are rebuttals, networking within the account, pre-call research, and so on. Never mind technology, list sources, lead sources, and other tools and all the questions they pose.
After considering all of these issues, it would appear that teleprospecting would be impossible to test. Yet every day, tests with high complexity are being conducted right before our very eyes, and the outcomes are readily apparent. In my next post, I will explore the details of this experimentation and how they can be applied to teleprospecting.
In the meantime, do you think you can test the teleprospecting function? If so, how would you go about it? What variables do you think need to be included? What are the most important questions to ask?