When you spend to acquire customers who only pay after a delayed sales cycle, measure that lag explicitly rather than hiding it inside one blended acquisition number that kills patient channels.
HubSpot
HubSpot is a B2B SaaS company, founded in Cambridge in June 2006 by Brian Halligan and Dharmesh Shah, that built the "inbound marketing" category and a subscription platform for filling and managing the customer funnel. Its stories share one through-line: because HubSpot sells recurring software through a staged inside-sales funnel, there is always a lag between when it spends to acquire and when it actually gets paid, and getting the customer economics right means measuring that gap rather than hiding it inside a single blended number.
Acquisition Economics: why cost-per-lead is not cost-per-customer
The problem. HubSpot runs an inside sales model where this month's marketing generates leads that do not become paying customers until they clear a sales cycle. On average it takes roughly 60 days for a lead to convert to a customer, so collapsing everything into one Customer Acquisition Cost hides that lag and risks killing a channel that looks expensive this month but pays back next quarter. CPA (cost to acquire a lead, trial, or activated user) and CAC (cost to acquire a paying subscriber) are different metrics that get dangerously confused.
The approach. HubSpot tracks CPA at each funnel stage, Cost Per Lead, Cost Per Sales Qualified Lead, and Cost Per Trial, and treats those as the leading indicators that feed the eventual CAC of a paying customer on a Basic, Pro, or Enterprise plan. It also read CAC against lifetime value channel by channel rather than in aggregate.
How it solved it. The per-channel view surfaced the fact that selling directly to very small businesses produced only about a 1.5 to 1 LTV-to-CAC ratio, while HubSpot's partner channel, selling to the same small businesses, produced roughly a 6 to 1 ratio. HubSpot leaned into the partner channel instead of direct selling to that segment, a decision that would have been invisible under a single blended CAC.
Recurring Revenue: a subscription base that had to stop leaking
The problem. HubSpot sells its platform on monthly, quarterly, and annual subscriptions, so its whole business depends on recurring revenue and on keeping customers longer than it costs to acquire them. When it filed its S-1 in August 2014, its annualized subscription dollar retention was only 88.6 percent, well below the 100 percent threshold that signals healthy SaaS unit economics, meaning the customer base was leaking revenue faster than existing accounts were expanding. Average subscription revenue per customer was about $8,478 for the three months ended March 31, 2014, a small-business price point that left little room for high churn.
The approach. HubSpot treated net revenue retention as the metric that governs a subscription business and worked to push it above 100 percent by expanding within existing accounts (more seats, more products, upgrades) faster than customers churned out, moving upmarket from the very small businesses that churned most.
How it solved it. Net revenue retention climbed from 88.6 percent at IPO to a peak around 115 percent, and the recurring-revenue engine scaled HubSpot to roughly $2.6 billion in revenue and a market value in the tens of billions, evidence that the SaaS model works only once retention crosses the line where expansion outruns churn.