Most marketers are obsessed with metrics, and this is particularly the case for online marketers. Metrics not only demonstrate performance, they often form the basis for decision making, creating buy-in and defusing politics. But, for the life of me I can’t recall ever seeing a decent presentation on metrics in any marketing book or course. In this post, I’m going to present a simple, process-centric approach to metrics that will help you develop the specific metrics that are right for your particular business and provide insight on where and when each metric makes a difference. This approach to metrics owes more to statistical process control (total quality) and activity-based costing than to marketing, but the the concepts are considerably (over)simplified. The goal is to achieve a comprehensive, integrated, insightful process view with as little math as possible.
Three simple concepts to master for developing great metrics
A process is a stable, repeatable set of dependent activities
Process = Activity 1 > Activity 2 > Activity 3 > … > Activity N
Activities consistently turn inputs into outputs at a certain efficiency and speed
Output = Input > Activity
Efficiency = Output Volume / Input Volume
Speed = 1 / Cycle Time
Activities are produced by resources that have a certain cost and capacity.
The bottleneck resource constrains overall output, therefore maximizing output
and minimizing costs amounts to improving throughput at bottleneck resources.
Total Cost of Output N = Resource1 Cost +…+ ResourceX Cost + … + ResourceN Cost
Unit Cost = Total Cost / Output Volume
Max Output N == Max Activity X == Capacity of Bottleneck ResouceX
Kinda theoretical right? OK, so here is a concrete example that I think will be useful to many readers. Below is an inbound new purchase process model for a typical B2B SaaS company.
General process metric model applied to a top-level B2B purchase process.
For a lower level process example, click here to see the model applied to email marketing metrics.
Chances are you’ve heard or seen most of these metrics before, but they were probably jumbled around on a spreadsheet in random order with a few critical metrics missing and few irrelevant ones thrown in for…uh…good measure. However, when you see them laid out as they relate to the purchase process, their meaning and inter-relationships become much clearer.
Tips for Defining and Using Great Metrics
Keep it simple
So, the reality is that since we are talking about sales and marketing, and not manufacturing nuts and bolts, our processes are never as clean and predictable as this little model appears. It will at best be an approximation to reality. But, you have control over that approximation. Since each process can be broken into smaller processes, and each activity can be broken into smaller activities, you get to choose the level of granularity. And, you get to choose the activities to include or not include. For example, the selling part of the process above is often broken down into excruciating detail, creating a lot of overhead reporting of sales activities that is never used to make a single decision. Define your process at a level that maximizes your understanding and control over your business, without weighing you down with meaningless reports. And, focus on the activities that have the biggest impact on output.
Focus on the bottlenecks
Most of the time there will be over-capacity or slack in one part of the process and under-capacity or a bottleneck in another part of the process. Continuous improvement is great, but it doesn’t do a bit of good to improve non-bottleneck activities, because you will always get stuck behind the bottleneck. For example, it won’t help to add more salespeople or shorten your sales cycle in an inbound sales model if you don’t have the leads to support the new capacity. You must increase the leads first. The right approach is to identify your current bottleneck, remove it, and then see where the next bottleneck appears, then remove it, and so on. If you’re interested in developing a better understanding bottlenecks, or going deeper on many of the ideas in this post, check out The Goal, a well known and easy-to-read novel (yes novel) on business process improvement.
Simplify by combining multiple metrics
Often there are many activities that combine to produce an output. Lead nurturing provides great example of this. You often don’t know which activities will work for which prospect, the optimal order of the activities or even which activities have actually happened. One way to keep your process and metrics simple and manageable is to roll up complex sets of activities into a single super-activity and create a metric for it that is a combination of the lower level metrics. This is the essence of lead scoring.
Visualize and measure the process from end-to-end
As they say, if you don’t know where you are going, then you can’t get lost. And, if all you have is a hammer, then everything looks like a nail. Well, you can easily get lost in a technology startup. And, if you waste your time hammering nails when you should be sawing wood, you can burn through a lot of funding before getting it right. Defining and measuring your sales and marketing process are simply a means to an end: making money! The real goal is to optimize it for maximum throughput. The process and metrics you have today can always improve. So, don’t settle for piecemeal metrics that don’t get you all the way to the finish line, measure the process from end-to-end. Moreover, don’t just use your process model to measure what is happening today. Know what you want it to look like a year from now, and use it to visualize the future and drive toward your goals.
Common Metric Mistakes to Avoid
If you don’t know your process, or don’t have a process, the first step is to define it and get it under control. You might not get it exactly right at first, but if you don’t have a repeatable stable process, then you won’t know what to measure, you won’t know what to do with the metrics you do measure, and you can’t consistently improve.
Forgetting the customer
In the final analysis, your sales and marketing activities will always be an incomplete set of the total set of activities that go into a sale. At best they play a supporting role to your customer’s purchase process activities. In an ideal world, you would define your entire process and metrics in terms of your customer, but in the real world you must usually settle for you own internally-focused approximations. As such, it is easy to start mistaking your own internal sales process for the real external purchase process, creating an illusion of control that does not exist. For example, in the process above an internal prospect record called a registration is used as an approximate measure of customer interest. But, just how interested is the prospect that registers on your website? Remembering will guide you as you make decisions about how to structure your registration process. How much work and information will you require of your prospect? Well, how interested do you want your prospects to be before you decide to spend more time and resources in the next step of the process?
Defining a fantasy process that ignores reality
A great example of this often occurs in enterprise sales. The marketing department will draw up a nice little process like the one above, where marketing leads turn smoothly into opportunities for sales. But, in reality 90% of sales opportunities are generated by sales activities such as cold calling, referrals and networking, not marketing leads. In this case, the overall process is driven by the number of sales reps, not the marketing spend.
There really is more than one process
Counter to my suggestions to keep things simple and combine multiple metrics etc., it is a mistake to combine truly independent processes into one. For example, it is usually the case that acquiring new customers is a distinctively different process from up-selling current customers. Some SaaS companies go so far as to separate the people doing the selling into sales reps and account managers. But, even if this approach is not right for your business, it is almost certainly the case that you should measure renewals, upgrades and up-sells separately from new business.
Choosing poor metrics
A good metric must correlate strongly to the real world idea you are trying to measure. And, it must be easily measurable. I mean, that is the reason you are using the metric instead of measuring the real thing. If the real thing were measurable, you would simply make that your metric, i.e., if you could really measure customer interest directly, then why bother counting registrations. You count registrations because You can. Alternatively, suppose you can also measure usage of a trial account too. It might be that you have lots of people signing up, but very few actually trying out your service. It might be that you can’t afford to spend time and money on look-y-loos and still turn a profit. In this case, you might be better off tracking trial usage instead of registration as your primary measure of interest.
When you calculate a unit cost, it is important to include ALL the costs of ALL the prior inputs and activities that went into creating that output. For example, the “cost per lead” generated by an email campaign to your nurturing database is not the cost of that campaign only. It is also all the costs that went into acquiring and maintaining the database all the way back to the original pay-per-click ad. And, your total acquisition cost should include all fully loaded sales and marketing costs. The easiest way to calculate total acquisition cost is to go straight to your accounting books, add up ALL sales and marketing costs including management, benefits, admin, etc., and divide it by the number of orders.
More posts in the SaaS Marketing Tips Series: