Manage the living s*** out of your pipeline: A how-to and sample model for B2B marketers

Dear B2B marketing peeps,

If you want to be good at your job but don’t model your pipeline within an inch of its life, measure the crap out of it, and refine it frequently, you’re doing it wrong!

This isn’t rocket science, but it does require some diligence. If you really get good at it, you can even get proactive (and, dare I say, predictive?) enough to smooth out your marketing-delivered pipeline when the stuff hits the fan (which never happens, of course!). Here are seven practical steps and a sample model.

Note: These are simplified versions of my models, but with the numbers changed and generic program names. I’d love to find a more elegant approach, either through more sophisticated modeling or a clever tool that won’t cost me an arm and a leg. If you have ideas, send them my way.

Step 1. Get in sync with your sales leader

Be totally in sync with your sales leader. Don’t even keep separate models! Work off of the same bookings requirements (whether over-assigned or straight) and agree on stages, stage definitions, and decision gates. Be on the same page when it comes to assumptions, including average selling prices (ASPs), conversion percentage and timing, and marketing vs. sales contribution. If you disagree on things, duke it out or bring it to the big boss. (On second thought, that will piss her off, so just work it out.)

Step 2. Build a bottoms-up model

Build a bottoms-up model for each product or segment. For this exercise, we will use these stages: 1. Lead. A suspect at the top of the funnel like a website visitor; 2. Marketing qualified lead (MQL). A lead that marketing believes meets your criteria to be a viable prospect based on demographics (role, industry, account size, implemented technology) and intent (they took some action like downloaded a white paper, attended a webinar, or filled out a contact form); 3. Sales accepted lead (SAL). A lead or contact that sales believes meets your criteria to be a viable prospect; and 4. Sales qualified lead (SQL). A lead that sales confirms is qualified to be considered an opportunity based on your criteria (typically because they have the budget, authority, and need to purchase your product, and plan to do so in a specified timeline (such as within the next six to 12 months).

Start with bookings and calculate how many deals you need based on your ASP assumptions.

Now that you know how many opportunities you need to have in each quarter, figure out how many opportunities you need to generate in each quarter. This is where your conversion timing assumptions come into play. I organize mine like this in a separate tab.

Go back to your original tab and, for each quarter, calculate how many SQLs you need to generate for that quarter and future quarters based on your SQL-to-close timing assumptions.

Add them up for each quarter at the bottom.

Take the percent of pipeline that you need to generate (tie off on this with your sales leader) and re-calculate the actual pipeline and SQLs you need to generate in each quarter.

Pro tip: What if you’re doing account-based-marketing (ABM)? Good on ya! Back out your top-of-the-triangle ABM accounts (the 100 or so you’re going to be all over each quarter, and assume a certain percent of those close within your time period). The better able you are to close those accounts, the lower the rest of the pipeline burden is on you. Use this model for everything else — lower-level ABMs, vertical targeting, and all other programs.

After you’ve calculated the SQLs that you need to generate in the quarter, build off of that number to calculate SALs, MQLs, and leads based on conversion and conversion timing assumptions. When you’re done calculating leads, that should help you understand whether you’re investing enough in marketing programs based on your budget and a quick-and-dirty calculation of your cost per lead (CPL). It should also foot with your estimated leads from the programs in your tops-down model. Here’s the quick-and-dirty CPL calculation.

Step 3. Build a tops-down model, too

It’s one thing to know how many SQLs, SALs, MQLs, and leads you need, but another to to know how many you’re getting. Take a tops-down approach to estimating how many leads you’re expecting to get from the programs you’re investing in. Here’s a simple view of top-of-funnel programs. In this example, the 2,050 leads in the first quarter do not foot with the 3,673 calculated leads in the bottoms-up model above, which tells you you need to plan for more programs. This should go without saying, but if you’re not yet confident in your programs, you should try to overshoot the mark.

If you have decent visibility into how your programs convert and you have predictable in-funnel conversion programs, you should do a secondary (and tertiary) model that helps you articulate the in-funnel leads you’ll be converting and how many leads or opportunities you think they’ll deliver. For example, if you have a predictable nurture program for converting leads to MQLs or a local event program for converting SALs to SQLs, you should create a similar model for each stage conversion.

Step 4. Measure performance regularly, and take action

Watch overall and campaign-level performance daily and report on it weekly. One of my colleagues calculates an expected run-rate for every stage so we know at any moment whether we’re ahead or behind, and by how much. Here’s a sample measurement for one of the stages. You should have drill-downs into a list of leads and opportunities from your overall dashboards as well as your campaign-level ones. Your drill down should include account (and whether a target account), industry vertical, company size, geo (to make sure you’re building pipeline in regions that make sense and for all of the sales regions if you’re organized that way), product interest if you have multiple products, and current status.

Step 5. Bring others into your orbit

Do not hoard this information (or any information, for that matter!). Share it far and wide within your company. Keep your executive team, especially your sales leader, apprised so they can see the sausage-making in action. This helps them understand what it takes to build pipeline, shines a light on issues that arise, keeps surprises to a minimum, and enables others to help you. It also helps you make the case for more program budget when you need it.

Step 6. Tweak your model regularly

Tweak your model at least once a quarter to account for all of the moving pieces, such as growing ASPs, conversion changes, and new product introductions. Don’t change it too often, though, so you know how you’re doing against a common target each week within the quarter. Make sure to document changes and keep old models so you can explain your thought process. Nothing’s worse than not having a good answer to this, especially in the board meeting!

Step 7. Get proactive, and maybe even predictive

By knowing what numbers you need, having an estimate for where you’re going to get them, and measuring progress against your estimates every week, you will have all of the pieces you need to be proactive. You’ll know when you’re on track, what programs are performing well, and begin to predict when you’re off the rails. As your campaigns stabilize (and they don’t always when you’re a startup, so segment them enough so you can get a read on mature campaigns while you continually layer on experimental ones), you know where you can swap in a better-performing campaign or double down on campaigns if you need more leads or conversions.

Want to download a sample version of the model (with dummy numbers, of course)? Here ya go.