Instrumenting the Machine

In the same way that we need instruments to understand how our car is running, we also need instrumentation to understand how well our Sales and Marketing Machine is running. Without measurement, there can be no improvement. The goal of instrumentation is to show us what is working well, and what is not working, so we can make adjustments to fine tune the process.

A machine also has levers that we can pull to adjust how it is running. In the car analogy, these are the accelerator pedal, brake, and steering wheel. For our Sales and Marketing Machine, these might be our spend on SEM (Search Engine Marketing) to create leadflow, the number of sales reps we hire, etc. We will also want to use the instrumentation to understand the effect of changing these levers, so we can optimize their settings.

What we want to know

Before we start, it is a good idea to think about what we really would like to know about our machine. Here are some thoughts on that topic:

  • We would like to be able to spot a shortfall in bookings far enough in advance to be able to take action and rectify the problem. Requires us work backwards from the steps that lead up to a closed deal. We need to check that the numbers of prospects in those stages are what they need to be, and that the Conversion Rates between stages are working as expected.
  • We would like to know what adustments are needed to hit growth targets for the following quarter/year. e.g. if we want to double revenue, what needs to happen to leads numbers, sales hiring, etc.
  • We would like to understand the cost of the entire sales and marketing process, and its various constituent pieces.
  • We would like to know which sources of leads have the highest ROI. To do that we need to know the cost per lead, and the conversion rates to closed deals for each different lead source. Looking at that data combined with the average deal size , will allow us to determine the return on investment for each different lead source.
  • We would like to know the effectiveness of the various Actions that we are taking to move prospects along the sales cycle. For example, if we are doing Webinars, how effective are those at getting attendees to sign up for a trial? Or if we are spending money on Pay per Click Google Adwords, how well are those visitors to the site converting into registered users on the site (where they have given their contact information)?

Spotting a bookings shortfall well in advance

Rather than applying instrumentation randomly to various points in the process, we should try to determine which parameters really matter. To do this, I recommend that you think back to the model that you use to forecast sales. What do you believe are the key drivers of sales growth? Is it the number of sales people that you have hired, and their productivity? Is it the number of resellers that you have signed up and made productive?

Start at the end and work backwards

I have found that the best way to understand how the model works and what instrumentation you need, is to work backwards from bookings. Bookings can be determined by looking at the number of deals that closed, and the average deal size. So we clearly need to instrument those two variables.

Now we need to go back a step, and look at what creates a deal. This will vary depending on our go-to-market channel. Here are some possible examples:

  • Direct Sales: bookings = the number of sales reps that are fully productive x the average productivity per rep x the average deal size

    Bookings could also be forecast by number of POCs (proof of concept) completed x conversion rate to closed deals x average deal size

  • Reseller Channel: bookings = no of mature resellers x average productivity of a mature reseller x average deal size (plus deals from less mature resellers, or non-certified resellers)
  • Touchless Web Sales for SaaS: bookings = number of trials x conversion rate at the end of the trial.
  • eCommerce: bookings = number of shopping carts x conversion rate (opposite of abandonment rate) x average basket value.

The above clearly shows which variables we will want to track to be able to understand what is happening to the key elements affecting bookings.

Next we will want to go back one step in the process, and look to see what comes before this final step. For example, this might be:

  • Direct Sales: sales hiring and number of POCs.
  • Reseller Channel: the number of resellers trained and on-boarded x the conversion rate to mature reseller status.
  • Touchless Web Sales for SaaS: number of visitors to the web site, and the conversion rate of visitors into trials.

To fill out the complete set of instrumentation, keep going back one step until you have reached the starting point. So for direct sales, before deals turn into POCs, they may be identified as Opportunities, and before they become Opportunities, they may be identified as Marketing Qualified Leads, etc. For each stage, you will want to track the Conversion Rate. e.g. how many Marketing Qualified Leads (MQLs) turn into Opportunities.

Instrumenting the effectiveness of Actions

As described elsewhere, the key building blocks or our Sales and Marketing Machine are Actions that we take to move the prospect through the sales process. For each Action, it is useful to know:

  • The cost of that Action per lead
  • The number of Actions that you completed. For example: how many visitors did you get to your site, how many attendees came to your webinar, how many trials took place, how many proof of concepts were completed.
  • The Conversion Rate that you got from taking that Action. The conversion rate will always be a percentage that measure the percent of leads that moved to the next step as a result of taking this Action.  For example what percentage of prospects that visited your web site converted into registered visitors; what percentage of webinar attendees signed up for a trial, what percentage of trials converted to a closed deal; etc.

These are best shown with trend lines showing how they are moving over time.

Be Sociable, Share!
  • http://www.opensourceadvisory.com Fred Holahan

    David, I've been doing quite a bit of work recently around instrumenting market engines, particularly from the demand creation point of view, so your post has some very useful insights. I agree completely that, once a critical few conversion metrics are known (or can be reasonably guessed), the best way to instrument the model is to work backwards from bookings to pipeline, pipeline to qualified leads, qualified leads to raw leads.

    I will add a few thoughts to the mix. One is that all leads are not equal in terms of the velocity with which they move into sales and the rate at which they convert to pipeline and deals. This is in part an ROI topic (as you raise) and in part an issue of making the operational model as accurate as possible.

    A second thought is that the volume model has to be pretty well designed and managed to deliver great value. For example, the act of assigning targets for different categories of leads should ripple through the model to keep reports, scorecards and charts in synch. Getting this right takes thought and effort, but it's hugely important.

    A final thought is that a well designed engine will often expose shocking insights. There are few things as alarming as having your own numbers tell you your demand generation forecast needs to be increased by a factor of 3 or 4 to hit a desired bookings target. It's cold, hard reality at its best.

    I have a post related to this topic at http://opensourceadvisory.com/wordpress/?p=1431. I hope you don't mind me posting the link, and look forward to more discussion on this in the future.