We said watch the transfer counts on Celo. They fell, and so did the volume

−70.1%

Weekly transfer volume

$1.62M to $482,915

−44.8%

Transfer count

839 to 463

6–7

Active addresses a day

the base under both moves

8.0%

Of cNGN transfer volume

down from 20.7% last week

Two weeks ago we reported cNGN's Celo deployment up 670%, and wrote that the thing to watch was whether transfer counts kept climbing while average size fell — because that, and not the volume figure, is what adoption looks like.

Last Monday counts had climbed eightfold and the average held flat. We called that half the signal, and said so.

This week both halves went the other way.

Volume −70.1%. Transfers −44.8%. Celo's share of cNGN transfer volume fell from 20.7% to 8.0%.

The base was always the story

The Celo deployment averages six to seven active addresses a day.

At that base, a 670% rise and a 70% fall are the same event described twice: a small number of participants moving different amounts in different weeks. Neither figure was ever evidence of a growing user base, and we should have led with the address count the first time rather than the percentage.

This is the failure mode we have written about before — a percentage on a thin base reads as a trend and is not one. We published the test in advance, which is the part that worked. What we would do differently is put the denominator in the headline.

What we still think

cNGN's move to Celo is real and the token is genuinely multi-chain: BSC carried $3,241,500 this week and Base $1,769,105, both on thousands of transfers rather than hundreds.

Celo is not yet part of that. It is a deployment with liquidity and almost no users, and until the daily active address count moves out of single digits we will report its percentages with the base attached or not at all.