Multishare
Back to Insights
Blog

Page Speed Is a Revenue Metric, Not an Engineering One

A 100ms improvement in load time is worth measurable revenue. Here is how to find yours.

The framing problem

Page speed lives in engineering backlogs, competing against features. It loses, because features have owners and speed has none.

Reframe it as revenue and the priority changes.

What the relationship looks like

Across e-commerce sites we have instrumented, conversion rate degrades non-linearly with load time. The damage is concentrated early:

LCP

Relative conversion rate

1.0s

index 100

2.0s

index 96

3.0s

index 87

4.0s

index 74

5.0s

index 61

6.0s+

index 48

Going from 5s to 4s recovers more revenue than going from 2s to 1s. If you are slow, the first second you cut is the most valuable one.

Calculating your number

Monthly sessions        × 250,000
Current conversion      × 1.8%
Average order value     × $42
                        = $189,000 monthly revenue

Move LCP 4.0s → 2.5s   → est. conversion 2.25%
                        = $236,250 monthly revenue
Delta                   = $47,250 / month

That is the number that belongs in the sprint planning conversation. Not "improve LCP."

Where the time usually goes

In our audits, the largest contributors in order of frequency:

  1. Third-party scripts — analytics, chat, A/B testing, ad tags. Frequently 40%+ of blocking time.
  2. Unoptimised images — wrong format, wrong dimensions, no lazy loading
  3. Render-blocking CSS — no critical-path extraction
  4. Server response time — no edge caching, uncached database queries
  5. Font loading — no font-display: swap, no preload

The uncomfortable finding is that the top item is usually marketing tooling. Every tag was added for a reason, and nobody owns removing them.

A pragmatic first pass

  • Audit every third-party script and remove anything without a named owner
  • Set a performance budget and enforce it in CI so regressions fail the build
  • Serve responsive images in modern formats
  • Preload the LCP element
  • Measure with field data (real users), not just lab tests

Lab data tells you what is possible. Field data tells you what your customers experience — and those diverge most on the mid-range Android devices that make up much of the Vietnamese market.

Explore web optimization services.