Days To Go Live

What it measures

For each customer with a completed activation in the period, count the calendar days from contract signature date to the date the customer first begins paying (or agreed start date, if that precedes first billing). Sum these durations and divide by the number of activated customers to get the average.

Why it matters

Days to Go Live is the operational heartbeat of customer success. Long activation windows tie up deployment resources, delay revenue recognition, increase buyer's remorse risk, and lower NPS. Shorter times free up resources, reduce friction, and set customers up for success faster. Watch it to diagnose whether implementation is a bottleneck to growth.

How to read it

Track DTGL month-over-month and quarter-over-quarter. A rising trend means your onboarding process is slowing — investigate root cause (team size, complexity, training gaps, technical blockers). A falling trend is a green flag: you're becoming more efficient. Compare your DTGL across customer segments (by size, industry, product) — if enterprise customers take 40 days and SMB take 8, it's a signal to either invest in enterprise playbooks or pre-sell with realistic expectations. Always segment by cohort: new customers often activate slower than expansion deals.

What good looks like

Watch-outs

• Including same-day or instant activations as zero days. Some customers may activate on signature date (e.g., self-serve signup); these are valid DTGL=0, but don't use them to mask slow implementations for others. Segment and report median *and* mean.
• Using different definitions of 'activation' across teams. Finance counts first billing; Product counts first feature access; Success counts 'customer ready to use.' Align all teams on a single definition (typically first billable date) before calculating.
• Ignoring customer size or type in aggregate reporting. Enterprise implementations naturally take longer; reporting a blended average hides whether SMB onboarding is degrading. Always break DTGL by segment (product tier, region, customer size, use case).
• Counting only 'successful' activations. If 5% of contracts never activate, exclude them and report only completed activations, or DTGL will artificially drop when you fail to onboard customers at all.

Worked example