Modern marketing comes down to just one thing:
Creating efficient revenue growth.
This is a bolder goal than many “traditional” marketers may subscribe to but I believe it is the right one. This means that marketing should have an eye on new sales, existing clients, and costs… and the analytics to tie it all togehter.
So here it is… the 3 biggest numbers to track:
- New ARR bookings (and all leading indicators)
- Sales funnel and pipeline health
- New sales opportunities created
- New sales-qualified leads
- Churn and up-sell (and all leading indicators)
- Customer/revenue churn
- Up-sell and cross-sell rate
- Customer satisfaction
- User engagement
- Customer Acquisition Cost (and associated efficiency metrics)
- Win rate
- Sales cycle length
- Revenue per lead
- Cost per lead
Each of these metrics can be endlessly broken down by channel, market segment, geography, etc., and those will help executives and team members keep track of pockets of strength and weakness within the overall business.
1. New ARR bookings:
Shouldn’t Sales be worrying about this? No. Everyone needs to worry about this.
More importantly, it’s critical that Marketing have an end-to-end view for what is and isn’t working. In a healthy organization, insights from some of these metrics can help inform sales leaders on how to coach and support individual reps. Other metrics will give great insights into where salespeople need better collateral, which leads need more nurturing, and which channels need more work (or more fuel).
However, you’ll need a crystal-clear and unanimous set of definitions before much progress can be made. Here are some suggestions, but this is something that will be different for every business:
lead ---------| opportunity-------------------------------------| client!
MQL > SAL > SQL > Demo > Selected as vendor > Out for signature > Won
- Marketing Qualified Lead: Marketing believes there is a person (or people) at the account that want to have a conversation with a sales rep. This should have even more specific criteria by channel: e.g., a webinar attendee requested a followup, an inbound demo request came in from a solid account, a late-funnel whitepaper download included name and contact details, etc. (track both individuals and accounts that make it to MQL)
- Sales Accepted Lead: Sales has attempted first contact with this lead. This is an important “in-between” stage that can help you measure the timing of outreach after leads are passed to sales. Once sales reps are following up with new leads in a matter of minutes then you may no longer need a “Sales Accepted” stage.
- Sales Qualified Lead: Sales has made contact with someone at the account and confirmed that there is indeed interest. At this stage, the salesperson should try to do lots of pre-demo discovery including identifying a full list of influencers/decision-makers, expected budget, buying trigger (why now?), other competitors in the mix, and top concerns… and then schedule an initial demo. (again, track this at both an individual and account level)
- Demo: I like to make this the first opportunity stage as it’s a good bar for “is there a possibility to make a sale.” To qualify as a Demo, the client must attend the demo and express a willingness to proceed with next steps. I recommend that marketers and sales managers sit on demos at least once a week to get a sense for what is happening at this stage.
- Selected as vendor: This is the second opportunity phase. I.e., the client has told the salesperson that they are the vendor of choice. Hopefully this means that there is an internal champion at this point. Urge sales reps to establish the “vendor of choice” before contract negotiation. This puts reps in a much better negotiating position.
- Out for signature: This means that the sales rep believes that favorable contract terms have been negotiated, decision-makers are bought-in, and the only thing that’s left is to sign on the dotted line. Of course, you may still have deals moving back to “Selected as Vendor” when a new set of stakeholders or legal requirements pop up after the first draft of the contract is sent over… that’s okay. Deals can move backwards.
- Won: that’s what it’s all about! At this point, Sales likely hands off to a Customer Success org. Make sure any transition work is done before letting sales reps move things into “won” … then go ring that gong!
Once you have a set of clear definitions, you’ll want to keep a watchful eye on how different reps’ pipelines progress. Ideally, differences in deal progression will come down to rep skill and not interpretation of the definitions. Investigating big miss-matches can be a great way to determine where reps need help from Marketing and/or sales leadership.
Based on measured or estimated progression rates from phase-to-phase you should be able to calculate what a “healthy” pipeline looks like and how long a deal should stay in each phase before alarm bells start to sound.
Furthermore, with clear definitions and predictable progression rates and stage-timing you can truly start to use lead volume and demo-volume as good leading indicators of revenue. This is the holy grail of predictable revenue: “we got X leads this week so we’ll probably have $Y in Z months.” Awesome.
You’ll probably want to assign a bunch of meta-data to each lead and deal. Things like geography and customer segment are fairly easy to positively identify (though not always). The one that I’ve found is the most tricky is channel. This topic has been well documented and there are a thousand different recommended approaches.
For simplicity and sanity I use “MQL-attainment-channel.” I define this as the last channel that touched a person before they became an MQL. So if a lead follows any of these three paths they count as “demo request”:
- Met at a trade show > clicked a nurturing email > attended a webinar > filled out a demo request form
- Got a cold-call from an SDR > downloaded a white paper > filled out a demo request form
- Downloaded a white paper > read everything on our blog > filled out a demo request form
This will do a good job of making “lead channel” a good proxy for lead quality. Here are the other lead channels and the MQL-attainment criteria I’ve used in the past (roughly in order of quality).
- Sales prospected – salesperson met the lead via their own networking activities?
- Web inbound – lead filled out a demo request on the website
- Inbound call – lead called or emailed and asked to speak with a sales rep
- CS Referral – lead was passed from the Customer Success team as a client referral
- SDR qualified – sales development reps qualified the lead and passed to sales
- Trade show – lead asked for a followup at a trade show
- Webinar – lead asked for a followup from a webinar
You’ll also want to track “how we first heard of this person” – I call this “lead source.” This isn’t exactly the same as “first touch” given that you may not know what someone’s “first touch” was. I’ll talk more about why this is important in section 3 on CAC.
2. Churn and up-sell
Shouldn’t Customer Success be worrying about this? No. Everyone needs to worry about this.
In a SaaS business keeping existing customers happy is critical. Growing existing accounts via up-sell and cross-sell should be supported by marketing, and marketing can also help with tools and tactics to keep customers happy and engaged. Bonus: really happy customers also tend to give more referrals, case studies, and references which helps with the numbers above.
A lot of SaaS companies like to define churn in all kinds of funny ways. When I say customer churn, I mean (how many customers did you lose) / (how many could you have possibly lost). Revenue churn should be measured in the same way. This gives a much more accurate representation of performance vs. adding in up-sell to create “net churn” or cheating with the denominator and including customers that can’t churn because they are still locked into a contract.
This is fairly easy to track, and the ultimate goal should be to keep this as close to zero as possible. Track churn by market segment, revenue-segment, geo, CS team, year-of-acquisition, etc. Every different angle you can inspect can give a hint as to where current or future problems may lie.
There are a few leading metrics for churn that you should also do your best to keep track of… these definition depends a lot on how the product is architected and how often you’re willing to survey your customers, so I’ll leave it a bit more open to interpretation.
- Company engagement – I like to think of this as how many positive touchpoints we have with the client. Opened a newsletter? Good! Had a “check-in” call with their Customer Success rep? Good! Posted a review? Good!
- Product engagement – How many active users do they have? How many features are they using? How much data do they have on the platform? All of these are good measures of stickiness. Try to distill this into a single numerical score.
- Customer satisfaction – you could also use NPS. The idea is to have a periodically measured metric for how happy customers are. Happy customers tend not to churn.
Okay, great, so you’ve got all the data tracked. What do you do with it? Figure out where soft spots are and address them! Newsletters targeted toward current customers should attempt to boost product and company engagement. Celebrate and surround satisfied customers to bring in referrals, references, and case studies. Work with Customer Success reps to find out what collateral and messaging they need to satisfy unhappy customers.
Up-sell and cross-sell
The other big opportunity with current customers is to make sure they are getting a steady drip of marketing information on new features, new products they can buy, and new ways to get more out of the product (especially if that comes with more ARR for the business). Depending on how structured your Customer Success team is, you can be as detailed in tracking new ARR from up-sell and cross-sell as you are in tracking new ARR bookings from Sales.
It may be enough, however, to simply track up-sell and cross-sell by segment and look for targeted ways to boost each of them with targeted marketing outreach, 1:1 CS interactions, or special offers.
3. Customer Acquisition Cost
So far, we’ve talked a lot about the revenue growth part… now we get to the efficient part. Since resources are finite you need to stay laser-focused on deploying them as efficiently as possible. Customer Acquisition Cost is easy to compute at a very high level: (total marketing and sales spend) / (new customers added).
The trick is to compute this at a deal-level, channel-level, function-level, etc. This allows you to get a really clear view of how your marketing efforts are performing and where you need to optimize.
I also look at a few other sub-metrics to get a more nuanced view:
- Total Marketing cost per acquisition – excludes the sales component of CAC
- Marketing program cost per acquisition – excludes marketing fixed costs (e.g., software, personnel, etc.)
- Total sales cost per acquisition – excludes marketing costs to show ROI for salesperson comp
While CAC is simple to understand and track at a high-level it quickly explodes into complexity when you try to get anything more useful or granular from it. Here are a few ideas on how to compute this simply and effectively.
First, assign all marketing costs to one of a few groups. The sum should be the sum of the ENTIRE marketing budget as well as any other costs from outside the marketing organization that could be cut should the company decide it never needed to add a new customer. Here’s a simple taxonomy:
- Big costs whose split is estimated by channel/geo/etc
- People – often a huge cost, split by how much time people spend on various tasks
- Office/HR costs – Finance will likely assign a portion of office cost to marketing
- Software – think CRM, Marketing Automation, Analytics, etc.; estimate a fair way to assign
- “Other” – anything not fit in somewhere else — try to keep this to less than 10% of the total
- Direct-billed MQL-generating costs assigned directly to channels and geos. This is spend that’s associated with direct generation of a new lead that passes to Sales and shouldn’t include “influencer” channels like advertising:
- Sales Development – include people’s salaries as well as software costs, list buys, etc. focused on generating new leads. If SDRs are also closing deals or qualifying leads you may need to assign some of their cost to the “sales” side of the CAC calculation
- Web development costs – contractors, hosting, etc.
- Trade shows – sponsorship costs, booth costs, travel, giveaways, etc.
- Self-hosted events – cost to sign up attendees, venue costs, travel, etc.
- Any other MQL-generating channels – depending on your business, this may include PPC, direct-mail, partner referral bounty programs, etc.
- Direct-billed “influence” costs which can also be assigned directly to channels and geos. These include activities and touches that are higher-funnel and are harder to track (i.e., print advertising)
- Advertising – print advertising, digital display advertising, etc. If you have the capability to determine who saw each ad, then it can make sense to break this up into smaller component parts
- Content generation – this may include blog posts, whitepapers, collateral production, etc. Largely things that may or may not affect buyers high-funnel and hence count as an influence to a sale
- Memberships & affiliations – all of the professional organizations that you need to sign up for to get listed some place or point to as part of a box-checking exercise
- Etc. – there will be specific cases for every company
Next, you’ll do the same with Sales costs. These are often WAY easier to split and track. It often comes down to:
- Sales salaries and commissions – OTEs are a good approximation if actual costs are hard to get
- Other office / HR related costs – phones, laptops, office space, benefits, etc. (split this as you do comp)
- Travel & expense costs by rep – if this can be pinned to a deal, great. Otherwise by rep is fine.
- “Other” – this can include sales training costs, presidents club expenses, etc.
Finally you can put this together into a deal-cost formula that allows you to pivot on CAC (or sub-CAC metric) by channel, geo, person, function, etc. allowing the granular focus necessary to optimize.
You’ll need to know the total number of deals that come in by channel (again, I use MQL-attainment channel), geo, sales rep, etc. This becomes the denominator of each of the CAC elements.
For each deal, the associated costs are as follows:
- Channel costs =
[total costs of channel in geo] / [number of deals in channel/geo]
- Influencer costs =
[total cost of marketing influencers in geo] / [number of deals in geo]
- Geo marketing costs =
[total "other" marketing costs in geo] / [number of deals in geo]
- Geo sales costs =
[total "other" sales costs in geo] / [number of deals in geo]
- Sales rep costs =
[sales rep comp] / [number of deals closed by that rep]
- Sales costs =
[sales costs directly associated with the deal if available (e.g., travel)]
This should give every deal a set of costs that can be sliced and diced as needed. Check that the total cost of all deals matches a total cost number that finance would recognize.
To make this work you must also pick an epoch over which to compute CAC and some corresponding assumptions. I like to assume: “spend this quarter links to deals this quarter.” While it’s not perfect, it’s simple and gives enough accuracy to make decisions.
Of course, this particular method has a bunch of issues. The fact is that even getting the above right is a huge challenge. For completeness, though, I’ll list a few limitations here. If you’ve got a way to address these, please let me know and I’d be happy to include them on this post.
- Sales cycle timing – “Some of the deals take 6 months to close” or pick some other number of months… this is a real concern. If spend 2 years ago actually generated the lead that closed today then our metrics are off a little. If you have perfect data from the past you might be able to do some gymnastics to address this. Otherwise, pick an epoch that makes sense (maybe it’s 2 quarters instead of 1 quarter) and just go with it.
- Some deals are harder than others – this is certainly true. A deal that a salesperson has to spend countless late nights negotiating “costs” more in terms of rep time than a deal that is signed with minimal headache. I’ve thought about including a “complexity” score for each deal and normalizing so that a rep’s time is more heavily weighted toward deals that take a lot of work.
- Not every lead sees all of my influencing material – also true. One lead may see one blog post and start a sales cycle while it takes another 2 years of constant bombardment with ads, emails, blog posts, and cold-calls to finally get moving. Some platforms track the digital side of this pretty well, but I still don’t trust it. If you had accurate data on how many touches people truly had online and off you could split the cost more fairly.
- All of my deals come at the end of the quarter – this is a big problem and makes charts of CAC very lumpy. If you do a monthly plot, you see huge CAC in month 1 (when very few deals come through) and tiny CAC in month 3 when you sprint to the finish. This is why I suggest a one-quarter epoch.
This is just a very high-level view of what metrics need to be tracked, computed, and understood by an effective go-to-market organization. All of these suggestions will be more or less useful depending on who your customers are, how the business runs, and a thousand other variables. I hope, however, that this provides a useful frame for thinking about the problem and a good place to get started.
Please let me know if you agree or disagree, what works for you, and any other resources you think could be useful!