August 2026 closed at $59,732 in revenue, across 156 sales. That is the biggest month of the year, past July's $57,884. Last August we did $60,908 across 126 sales, so the best month of the year is down about 2% on the same month a year ago.
That is not a surprise. Here is what I said in last month's update, word for word: "August last year was $60,908, the strongest month of that year. The easy comparison I got in July flips completely in August. If next month's year-on-year number looks bad, that's why." Here it is, and that is why.
The part I did not see coming.
156 sales against 126 is 24% more sales, for 2% less money. The average sale went from $487 to $383, down about a fifth. More customers paying us, less money arriving.
One line explains the whole month. New sales last August were $21,867. New sales this August were $1,066. That is $20,800 gone out of one category, and the total year-on-year gap is only $1,176. So everything else in this business grew by about $19,500, and new business handed all of it back. Renewals were up $19,089. Upgrades were up $2,384. Existing customers did their job. Nobody new showed up.
I said something close to this last month about July, when new business was $1,152. I called it the line that matters most to me, and the one that went the wrong way. That is two months in a row now. One month is a bad month. Two is a pattern, and the pattern is mine to fix.
This company grew everywhere except the one place that brings in people who have never heard of us.
Two more things about the shape of August.
- It is concentrated, more so than July. One transaction was $19,810, which is 33% of the month in a single line. The top five were 67% of it.
- The hosting split moved hard. Cloud revenue was up about $9,700 on 145 transactions against 102 last year. Data Center revenue was down about $10,900 on 11 transactions against 23. Half as many Data Center sales as last August. Cloud is where customers are going, and that is the direction we have been building in, so I am not unhappy about it. But Data Center is still where the large annual amounts come from, and I do not want to lose that quietly.
Refunds were $2,489, against nothing last August.
The number that actually measures the 10x.
The monthly figure is the drama. The trailing 12 months is the truth. After July it was $562,794. After August it is $561,617. It went backwards, by about $1,200, and it is still about 10.8% of the $5.19M target.
This number has gone backwards before, and harder. In May it dropped by more than $19,000, and it dropped again in June. What is different this time is the reason. May went backwards because May was a bad month, $13,987. August went backwards off the back of the best month of the year. Our best month of the year, and we are further from the goal than we were before it. That is what it looks like when you replace a strong month with a slightly weaker one.
Five months in, about a tenth of the way there, and nineteen months left.
The scorecard.
Against the three things I committed to for August:
- Release the advanced Document Vault for Jira tier. Not released in August.
- Release Secure Admin for Jira Cloud. Not released in August.
- Start prototyping two new Jira apps. Both started.
One out of three. But the real state of the two that missed has changed since the end of the month. Both of them are finished. Both are on the Atlassian Marketplace right now, waiting for approval. Last month I told you they were in testing. They are out of testing, they are submitted, and they are out of my hands. I am hoping both are approved by the end of September, and I want to be careful with that word. The timing belongs to Atlassian, not to me, and I am not going to promise a date that somebody else controls.
QR Studio is in the same position. In August it was rejected by Apple over how our purchasing tiers were set up. That is fixed, and it is back on the App Store waiting for approval. Three products finished, all three sitting in somebody else's queue. That is a better problem than the one I had last month. It is still not revenue.
Three other things.
We retired an app. STM Lite is being withdrawn rather than converted to Forge, and the three real customers on it were offered a move to STM Issue Templates in August.
SOC 2 moved. We are writing the documentation now. It is being created and reviewed. That is still not a certification, there is still no date, and I am not going to invent one.
And the two levers I keep reporting as untouched are untouched again. Nothing on the partner channel. Nothing on direct outreach to the compliance and security people. At some point I have to stop calling that a gap and start calling it a choice.
What is next.
- Get the advanced Document Vault for Jira tier and Secure Admin for Jira Cloud approved and released, with the caveat above.
- Start Comment History Advanced. That is the second app to get a tier, and the pricing lever spreading across the range instead of sitting on one product.
The next two months.
I am away from 1 October until the middle of November. I am going to Team 26 Europe, the Atlassian conference, and then taking a holiday.
The monthly updates keep going. Month 06 on 14 October and Month 07 on 11 November, both in the usual slot. The weekly videos stop until I am back, but the posts continue. November's update might be shorter, and it might just be my voice over footage from the conference. I would rather give you something small and honest than skip a month. I am not going to pretend six weeks away costs nothing. It does. But the team keeps working, the apps keep moving through approval, and I would rather tell you now than have you work it out.
So. August. The biggest month of the year, down year on year, and the trailing twelve months went backwards off the back of the best month of the year. This company grew everywhere except the one place that brings in people who have never heard of us. Either I pull this off or I don't. I will see you next month.