In a conversation with Supra’s Marc Baselga and Ben Erez (podcast here), we touched on a core thesis of mine that I’d never written in long form...
IMHO a company’s price point shapes the organization itself, including some fundamental C-level conflicts. An enterprise software company with $150,000 average sale looks and behaves very differently from a music stream service or consumer finance app charging $8/month. Staffing, incentives, escalations, experiments, power dynamics in the C-suite… all very different. That’s why I see B2C product managers (and their organizational assumptions) in contrast with B2B/enterprise product managers (and their organizational assumptions).
First, some basic math, borrowing a chart from Christoph Janz (at top):
- An enterprise company getting $100k/customer/year needs 10 customers (elephants) to hit $1M ARR, or 1000 customers to hit $100M ARR.
- A mass consumer company with $10 annual subscriptions needs 100k customers (flies) to hit $1M ARR, or 10M customers to hit $100M ARR.
Why Does This This Matters?
If we’re a mass consumer company at $10 ARR/subscriber, we might need to sign up another 5M users this quarter. That suggests massive outreach (marketing, influencers, referrals, automated funnels) and trivially easy demo/trial/signup/upsell and brilliantly effortless UX. And if we’re in a freemium model, we may need to add 100M free users to convert 5M to paid subscriptions. It’s all about scale, reduced friction… very large numbers of very small amounts.
Most of the writing about product-led growth (and product management in general) is focused on B2C: how to get huge numbers of individuals through trial or signup or freemium flows – where the product is the experience, great UX/design triumphs, and users are app-guided through each step to minimize human intervention or assistance. At $10/year, we can’t afford to talk with each person.
- There’s no classic sales organization in this model. (Or a small sales group focused on partners, suppliers, and channels relevant to our mass audience.) Marketing rules the roost, product/design/engineering delivers the goods.
- Huge numbers of users/interactions/signups/churned accounts let us do truly quantifiable experiments. If I have 90k visitors per day to the “choose your favorite artists” flow of my music streaming app, I can reasonably test alternate designs in an hour. We might be running 50 simultaneous A/B tests each day. This is decision science, as popularized by the FAANGs.
- The user is almost always the decider and the buyer. “Do I think that outfit will look good on me?” “What do I want delivered for lunch?” “Swipe right or left?”
- We need prospects to make decisions quickly/easily. If an individual has to come back several times, or weigh complex choices, or make long non-cancellable commitments, we probably lose them. “NOW” means 2 minutes or less, ideally punctuated by a couple of WOW moments or solid proof points.
All of that is important because there’s no one customer (at $10-$200/year) who has the CEO’s phone number and attention. At the executive level, we deal in aggregates: activations and upsells and churn in units of hundreds-of-thousands-per-month. We naturally look at averages and mass motion. So it’s very rare for the CEO to buttonhole the CPO/CTO/Head of Design with a single user’s complaint. (“I just talked with Sigrid Singersdotter, who is having trouble finding her father’s music with our search function.”)
On the Enterprise Side…
If we’re an enterprise software company, these are reversed:
- Sales is usually the most powerful (and best paid) organization. 5-10 major opportunities make up most of this quarter’s revenue, so the CEO and Board carefully track every one of those deals every week, and push the CRO for ways to help close – including CEO calls to individual prospects, who typically have an ask or two or three. The weekly exec meeting starts with a deal-level pipeline review. Salespeople who don’t close their top deals get fired. And if a couple of whales fall out of pipeline, 20% of our employees might be on the street.
- 9-to-18-month sales cycles makes feature-level attribution for new deals nearly impossible. Any individual (huge) prospect might get 40 drip marketing touches and 12 sales meetings and 3 demos and 60 website visits and 5 reference calls and 2 architecture/security reviews and 3 contract negotiation sessions scattered across a full calendar year. With <100 prospects/year, the numbers are too small to prove that any particular product feature or marketing message or sales technique or third-party integration or webinar or Gartner reference or CEO call was what tipped a prospect into the ‘buy’ column. (In the real world, it’s always some complex mix of many factors, even though each department takes the credit for itself.) This is data-assisted intuition with enormous error bars: it’s rare for any one improvement to dominate, and even rarer to get exec-level consensus about it.
- Buying committees and RFPs and C-level customer-side politics are central to enterprise purchase decisions. And these players tend to be far from the actual users of actual software systems. In my experience, most of the buyer-level demands that hold us hostage are unimportant, or are technically incorrect, or propose a wrong solution, or simply don’t reflect what’s actually important to their actual users. Anecdotes and out-of-context requirements are rampant, but there’s tremendous pressure for us (as vendors) to agree.
- 'Product-led' is important in designing each step of customer onboarding, activation/usage, support, merchandizing of value, and good customer experiences. And important for renewals. But enterprise products don’t 'sell themselves' and rarely generate enough of a groundswell from actual users to drive buying decisions. Employees use the software that leaders choose, even if those leaders will never put finger to keyboard. So our maker/product teams try to be product-led while the rest of the company is sales-led or investor-led or top-three-customer-recency-bias-led.
Which means that enterprise product folks see certain C-level behaviors much more often than their B2C counterparts:
- Pipelines are lumpy, and the future is uncertain. So current deals tend to overwhelm company strategy. (“I know that we’re focused on North American commercial banks, but this big UK buy-now-pay-later provider is asking how we can refit our product for their needs…”)
- Sales teams know that they have more clout than product/engineering. So a ‘no’ from product manager is the start of a negotiation, not a final answer. We hire and train and reward salespeople to get to ‘yes,’ not to meekly accept defeat and walk away. (FYI, a 'millisecond' is the time between when Product rejects a request and Sales escalates it to the CEO.)
- Single-customer interrupts happen every week. Any problem at a top ten customer goes right up the org chart, and displaces planned work. And we tend to universalize rare instances: “JPMC has a multi-factor authentication problem” quickly becomes “none of our customers can log in.”
- Roadmaps have initiatives/projects named for single customers. (“We’ve committed to finishing the Daimler/BMC integration by Oct 15th. Everything else is lower priority, including our next major release and our scalability/architecture work.”) One-offs and specials and obscure customizations are scattered throughout the code base, with their original developers long gone from our company. We don’t have a complete list of specials in production. Old things break all the time.
- Win/loss analysis can include a lot of blame-shifting. I’ve interviewed scores of enterprise sales teams over the years… and am never surprised to hear from them that we closed our major deals this quarter because of great selling. (“Our sales org is way above average.”) Conversely, Sales is strongly incented to attribute lost deals to missing features or pricing issues or late delivery or poor R&D productivity or other things outside Sales' immediate control.
Note that these are cultural and behavioral challenges, not analytical challenges.
Sales teams escalate to the CEO because it works. Immediate revenue trumps next year-strategy because we're afraid we may not survive until next quarter. (“I'll gladly pay you Tuesday for a hamburger today.”) Customers hold us hostage over desired features because vendors generally cave. CEOs insist on committing to 120% development utilization because it’s urgent just this one time.
Lectures about development processes don’t help. I’ve never seen a spreadsheet or a Jira chart convince execs to be more product-led (or more ‘strategic’.) Instead, I look for arguments that address their specific motivations and concerns: why investors value SaaS subscriptions at 10x more than one-time project work. Why Product X is likely to bring in much more revenue than Product Y. Identifying that we wasted $40M on work that was delivered on time but didn’t meet market estimates or sales projections. Explaining why successfully migrating 60% of our customers this year is better than 100% of them in 2031. With market-side leaders, we need to speak the language of money.
Sound Byte
Enterprise companies tend to have different product/market challenges than mass consumer companies, and therefore are staffed and structured differently. Decision-making is different. Escalations are different. So enterprise product leaders need an expanded set of political/strategic/communications tools to manage their C-suite. CPOs and Product VPs should remember that prioritization is a political problem as much as an analytical problem.