How ARR works
ARR is the recurring portion of subscription revenue expressed on an annual basis. A customer paying 200 dollars a month contributes 2,400 dollars of ARR. It excludes one-off fees and counts only the predictable, recurring component that defines a subscription business. Monthly recurring revenue, or MRR, is the same idea on a monthly cadence, and ARR is essentially MRR times twelve. For a SaaS business, ARR is the headline health metric, but for advertisers its importance is what it says about the worth of each new customer the ads bring in.
Why ARR matters for bidding
Smart Bidding optimises to the conversion value you give it. If a SaaS company passes the first invoice, say one month of subscription, as the conversion value, the algorithm believes a customer is worth a fraction of their true value and bids far too conservatively. Passing a value grounded in ARR, or better yet in expected lifetime value derived from ARR and retention, tells bidding what a customer is actually worth. That single change often unlocks the willingness to pay for the high-intent clicks that acquire durable subscribers.
ARR, retention, and lifetime value
ARR is the starting point, not the final answer, for conversion value. A subscriber worth 2,400 dollars of ARR who stays for three years is worth far more than one who churns in two months, so retention has to be layered in. The honest value to bid toward is expected lifetime value, which combines ARR with realistic retention and gross margin. Using raw ARR is already a huge improvement over first-order value, and refining it toward margin-adjusted lifetime value is the next step for a measurement-mature SaaS account.
How iClick uses ARR
iClick reframes SaaS conversion value around ARR rather than first-order payment, so Smart Bidding stops undervaluing durable subscribers. The method is to pass a value grounded in ARR and, where retention data supports it, in margin-adjusted lifetime value, so bidding is willing to compete for the high-intent clicks that acquire lasting customers. The rule is that the value passed to the algorithm must reflect what a customer is really worth over time, not what they paid on day one.
