Part 4 of this series focused on the possibility of industrializing the top of the sales funnel. The question has been raised: “Why just the top of the funnel?” After all, if we don’t close more deals and dollars, have we accomplished anything? What a great question!
Here’s my answer: unless and until we industrialize the top of the funnel, there is no point in even trying to industrialize the rest — from discovery through closing the deal.
Why? Because it turns out that for almost every B2B company, the top of the funnel is the constraint of the entire business. Do the thought experiment of doubling the flow rate of discovery meetings — the output of the top of the funnel — and you get a linear increase in deals and revenue.
Now, let’s do the opposite experiment — the one where the flow of discovery meetings slows as the easy inbounds and low-hanging fruit dry up — and the result is even more predictable. Fewer discovery meetings per day causes deal activity to shrink down to a desperate attack on renewals and bluebirds. This is the exact opposite of industrialization. Sadly, it is very popular, as diets of junk food that give way to starvation often are. The dreaded “knee of the S-curve” is inevitable, even for the fastest growing, hot tech companies, and when it comes, the withdrawal symptoms can be devastating.
One is the “only-est” number
The point of all this thinking and experimenting is to point out that industrializing the top of the funnel is an attack on the growth constraint of most B2B companies. Not a growth constraint, but the growth constraint. This is just math, but it’s math that bothers many people who react immediately by saying, “No, Chris, that’s only one of the many growth constraints on my business.” It turns out that in any system, there can be only one throughput constraint at a time. Of course, as one constraint is addressed and its throughput increased, it is possible (in fact, likely) that the constraint will move either upstream or downstream. But right now, there can only be one, and if we aim to industrialize sales, we are wise to start with that one. We can deal with the consequences of success later.
So, let’s address the residual questions left after Part 4 of this series. What about costs? And how about scaling? After all, if all this industrialization comes at too high a cost, we might be better off with our current cottage-industry approach, where each sales rep is a “factory of one,” and we manage by a combination of hiring, assigning territories, setting compensation plans, and firing when disappointed. And even if the cost is reasonable, if we can scale, why bother? We can chew through the immediate market opportunity in front of us with reasonable efficiency, but we have little chance of taking over the world, so to speak, if we can’t scale up sales to match our presumed ability to scale up delivery.
A little recap and vocabulary building
Here are the cost elements from Part 4 of this series:
- The cost of the data needed to support a unit of output (i.e., a conversation)
- The fully burdened cost of the labor needed to produce a unit of output (again, the conversation)
- The all-in cost of running a conversation machine to produce a unit of output (still, a conversation)
Unfortunately, conversations must be manufactured through a stochastic, or random, process in which we try something and respond to what happens as a result of that trial. It turns out that all manufacturing processes are stochastic, so this is not an impediment to industrialization. It just tells us that we need to think in terms of yields and their variability.
How to fill your data tank
What yields come into play for top-of-funnel sales? The connects-per-dial yield comes first. If we know what a dial costs, and we know the connects-per-dial yield, we can calculate our average cost per conversation.
First, let’s dispose of the cost and availability of data — that is, lists of potential target customers with enough information to support having conversations with them. The internet has radically changed the economics of data and continues to do so. In seconds, you can turn a question into a rich list of contacts. Business characteristics (cutely called “firmographics”) and people characteristics can be mixed and matched at will. Head over to ZoomInfo or DiscoverOrg, and you can fill your data tank to overflowing at costs that can’t possibly be a deterrent to fueling any real business. You can even automate an ongoing flow of companies and contacts that make sense as top-of-funnel inputs from ambitious companies like Node.io, which is “Google-izing” (Is there a correct spelling of this word?) the B2B web and dynamically mining it for your next best prospect.
The beauty of data obtained by the simple process of buying it is obvious but often overlooked. There is no S-curve lurking out there, or if there is, at least the S-curve is driven by your actual total available market size, not your total available “hand-raisers” rate. When yields on purchased data from these sources start to drop, they do so slowly, because the pool of data is vast.
Once again, it’s about the conversations
From an industrialization standpoint, it’s enough to know that data is available, and all we have to do is process that data by having conversations with the people it lets us speak with. So, let’s move on to where the going gets tough and where too many modern businesses turn back in confusion and fear. By that I mean using the telephone to call prospects you don’t know — the horrifying and despised world of the cold call.
Sure, we could send emails and leave voicemails, but eventually we need to have conversations. Cold calling (along with its close relatives — follow-up and referral calling) is the most direct way of getting targeted conversations. After all, you control the targeting (the list) and merely need to invoke the mechanism (the call). At some yield per dial, you will have targeted conversations. And those conversations will flow at some predictable rate per day. All of which will happen at some total cost per conversation. Remember, we are trying to industrialize here, and a simple design with predictable input flows obtained at known costs is essential to industrialization.
Alternatively, a series of “campaigns,” each one of which is an adventure that generates an unpredictable number of conversations at an unpredictable cost, is a lousy way to industrialize. You don’t see factory managers running “sourcing campaigns” or “purchasing blitzes” very often. Industrialization requires feeding the beast every day, and a series of adventures, fun though they might be, will yield a series of surprises. Some will be positive, and you won’t have the capacity to take advantage of their unexpected goodness; some will be negative, and you will starve while waiting for precious inputs.
So, let’s just make a flippin’ list that matches our ideal customer profile and dial the darned thing. If the math works, we have solved steps 1 and 2 on our path to industrialization.
What does a dial cost?
But, you ask, can we afford to dial? After all, “Nobody answers the phone anymore.” This is easy to test. All we have to look at is dial yields (conversations per dial) and, by applying dial labor and other costs, dial costs (dollars per conversation) to get an answer.
Typical dial yields range from 10% down to 2.5%; and typical dial costs range from $6.00 (figured at $300 per day of fully burdened sales development labor divided by a sustained 50 dials per day, which is roughly the industry standard rate) to well below $1.00 at a sustained 800 dials-per-day pace using an advanced conversation acceleration engine like ConnectAndSell Lightning™.
Fortunately, from an industrialization perspective, the actual process yield of conversations per dial doesn’t have much to do with whether we can industrialize. When combined with labor rates for actually having the conversations, it becomes important for any given attempt to industrialize. Why? Because our cost per conversation will be a primary driver of our cost per meeting. And our cost per meeting is a primary driver of our cost to acquire a new customer (for growth companies) or a deal with a current customer. And if that whole equation doesn’t yield enough of something we care about (future gross profit for regular businesses, or anticipated valuation increase for VC-funded startups — otherwise known as “R&D labs for unnamed future acquirers”), then we are screwed.
Radically lowering the cost of a conversation while holding quality constant is a clear formula for clearing this “Is our business viable?” threshold. In the above example, we have reduced our cost of acquiring a customer by a factor of 6 by doing only one thing: increasing the number of dials per day (remembering that these customers are not physically local, so we need to have conversations through the magic of telephony!) and amortizing our top-of-funnel sales labor across many more conversations. We probably can get above the threshold of viability at a paltry, industry-standard 50 dials per day per rep; but at 800 dials per day, our ability to acquire customers cheaply enough and fast enough increases non-linearly. (Remember, we always have overhead eating our capital while we sleep, kind of like owning a racehorse or a boat.)
So, let’s tuck the conversation yield and the conversation cost into bed and let them rest. Clearly, for any currently viable B2B business, if we can have enough targeted conversations at a total cost of about $1.00 each, we can take one step to industrialization.
Choose carefully, then coach assiduously
Variability is much harder to manage than yield. If each of our conversation machines (our reps) uses a different process inside the conversation, we have very little hope of controlling variability. All we can do is put a rep in place and wait to see how many meetings that rep produces relative to either a standard or, better, a baseline established from our own experience. Sadly, this is where it gets tough. Our means of selecting from among candidate top-of-funnel reps includes reading resumes (self-described), asking about prior performance (self-reported), checking a couple of references (self-curated), and — according to the best research — selecting a “mini-me” who, because they are “just like us,” must be just right.
However, it turns out we can industrialize the acquisition of our top-of-funnel reps who will transform conversations into meetings, referrals, follow-ups, and rejects. Here’s what we can do:
- Test them for characteristics proven to correlate to top-of-funnel conversational success. For example, the Objective Management Group assessment can be set up to reliably exclude no-hopers from the candidate list.
- Work them out, like an NFL team works out a candidate running back, before adding them to the team. Listening to 20 live conversations, as a candidate rep tried to sell your prospects on donating to the rep’s favorite charity, should be enough to exclude any duds who slip by the test.
There will still be variability, but it can be minimized by regular adjustment, called “coaching” in the world of sales. If a rep is executing more than 20 conversation cycles per day, this is a straightforward matter of monitoring yield, inspecting conversations both at random and when there is a statistically meaningful change, whether positive (because it’s good to detect and harvest possible spontaneous improvements for possible inclusion in the standard conversational process) or negative (because reps often “drift” into practices that feel good to them but radically decrease throughput).
One note: coaching is not a form of management. It is a separate industrial activity best thought of as “tuning the conversational machines to maintain throughput rates at established quality standards,” which is very different from “deciding if a conversational machine needs to be upgraded or replaced,” or for that matter, “predicting future output.”
What will you net?
The only remaining yield question to consider is “Will all those conversations predictably produce enough meetings to feed the next step in our sales process?”
This is, of course, the whole point of the industrialization exercise: to consistently produce meetings at sufficient quantity and of sufficient quality to run an industrial top-of-funnel sales process. The answer here is oddly simple: Yes. Why? Because it is a trivial matter to script the beginning of a conversation and, if necessary, adapt that script to input lists with different characteristics. (For example, sales leaders would get one script, and marketing leaders would get another.) It is also straightforward, although not trivial, to train conversational machines to consistently respond to a known set of additional inputs, usually called “objections.” These inputs can be reliably ascertained by analyzing conversations to generate playbooks, or sets of responses, that maximize the probability of the conversation leading to a meeting.
By these simple steps, meeting yield and, therefore, meeting cost and flow rates can be brought above the threshold required to produce net new meetings and, therefore (after additional processing), net new customers with known characteristics. Given the assumption that almost all B2B companies are bottlenecked on the production of quality meetings in sufficient quantity to fulfill their business plan, we don’t need to explore the question of whether these meetings will yield new customers and deals in a reliable way. We hope for that to become the bottleneck because of our ability to scale production of meetings through an industrial process; but until it is the bottleneck, we are wise to leave it alone.
Try it! You’ll like it!
So, in summary, by focusing our attention on predictably generating conversations from inputs that we can practically obtain — mere lists that derive directly from analysis of our strategy — we see that we can industrialize sales, at least at the top of the funnel. It is not easy, nor is it “one and done.” Like any good industrial process, it requires continual monitoring (of inputs, processors, and processes) and adjustment in order to be reliable both today and tomorrow. The good news is that the production of conversations itself has already been achieved and can be experienced at no cost.
There is massive competitive advantage to be had for any company that succeeds in industrializing sales while their rivals continue with the “cottage industry” model of sales, which makes exploration of this possibility not only exciting but compelling. And, given the dire consequences of having your most able competitor choose to industrialize while you continue with the ancient tradition of “hire a rep and turn ’em loose on a territory,” exploring the question through action rather than mere reading might be more urgent than you think.