The day the market asked what your tools are really worth

A quiet thread runs through today's announcements: the question of what you can actually verify about the software you pay for, and the money moving beneath it. From the economics of mature software to the funding that keeps open tools alive, the reporting keeps circling back to a single practical test — can you check what you got.

That test matters more than any feature list. A tool is a bet on a vendor, a bet on the money behind it, and a bet on your own ability to see whether the bet is paying off. Today's items are useful precisely because they touch all three.

The economics underneath the software you buy

SaaStr published a piece on what it calls the terminal state of software, and the argument is worth understanding before you sign a multi-year contract [1]. For two decades, the reliable business model in B2B software was the slow annuity: modest growth, but revenue that renewed year after year because your data was locked inside the application and could not easily leave. As the piece puts it, "Growth faded, but the annuity was forever" [1].

Why does this matter to a buyer rather than an investor. Because the model that made your data hard to move was never designed for your benefit. It was the mechanism that made the annuity durable. When you evaluate a tool, the honest question is not only what it does today, but whether you could leave it with your records intact. That is the difference between a supplier and a trap. We have written before about the plain version of this test — the data you should be able to export on any Tuesday — and it applies to every vendor, regardless of how the market values their category this quarter.

Who pays to keep open source alive

GitHub reported that the community has now contributed $100 million to the people who build and maintain open source software [2]. The company framed it as "$100 million contributed by the community to the people who build and sustain open source every day" [2].

Most businesses never think about this, and that is exactly the risk. The tools you buy are built on layers of open code maintained by people who are often unpaid. When a maintainer burns out or walks away, a dependency you have never heard of can stop receiving security fixes, and the commercial product sitting on top of it inherits the problem. Funding of this kind does not remove that risk, but it is a signal about the health of the foundations. When you assess a vendor, it is fair to ask what they depend on and whether that layer is maintained. A product is only as sound as the parts it does not advertise.

When your finance tools carry money between them

Xero announced a partnership with Revolut Business aimed at helping New Zealand firms trade internationally [3]. The framing is about ambition and reach: "Because our market is relatively small, many Kiwi small businesses have the goal to go global right from the start" [3].

The detail worth extracting is not the geography. It is that accounting software and a business banking service are being connected directly. Every such connection removes a manual step — and every manual step you remove is a place where numbers used to drift out of agreement. When your ledger and your bank do not talk, someone re-keys figures, and re-keyed figures are where errors live. This is the same principle behind why your tools do not talk to each other: an integration is not a convenience, it is the difference between one version of the truth and several. If you operate across currencies, a link between the tool that records money and the tool that moves it is worth more than most headline features.

Charts that show more than one thing at once

Google Sheets added support for combo charts, which let you plot different kinds of series together — for example, bars and a line — on the same visual [4]. Google described it as "enhanced support for combo charts" for "creating multi-series visualizations" [4].

This sounds small, and as a feature it is. As a habit it is not. A combo chart is the right tool when two measures move on different scales and you need to see them in relation — revenue as bars against conversion rate as a line, say. Used well, it makes a relationship visible that a table hides. Used carelessly, it invites the two most common chart mistakes: pairing series that have no real connection, and choosing axes that flatter a trend. A chart is an argument, and a combo chart is a stronger argument than most, which means it deserves more scrutiny, not less. We covered the discipline of reading your own numbers honestly in reading your website analytics without fooling yourself, and the same caution applies the moment you put two lines on one grid.

What a media buy actually delivers

SaaStr also published an unusually direct account of one of its own advertising deals: 5.4 million impressions that drove 11,749 people to a signup form [5]. The piece opens by naming the problem most buyers know and few say aloud — that with a typical media buy, "you sign up, wire the money, the ad 'runs,' and 60 days later you get a PDF with an impressions number and no way to check any of it" [5].

Set aside whether these particular numbers are good. The lesson for a buyer is about what you can measure, not what you were told. An impressions figure is an input. A signup is closer to an outcome. The gap between the two is where most advertising money quietly disappears, because impressions are easy to report and hard to challenge. When you spend on distribution — of any kind, not only media — insist on a metric you can trace to something real. The right question is never how many people saw it. It is how many did the thing you were paying for.

The through-line

Five announcements, one recurring demand: proof you can inspect. The economics of software reward vendors who make your data hard to move [1]. The tools you rely on stand on foundations that need funding to survive [2]. The value of connecting two systems is that the numbers can no longer disagree [3]. A chart is only as honest as its axes [4]. And a marketing spend is only worth what you can trace to a real result [5].

None of this requires you to distrust every supplier. It requires you to keep asking the same plain question — can I check this — and to prefer the tools and partners whose answer is yes. That is the whole of good buying, and it is the reason we keep returning to how to choose software worth using. The market's mood changes weekly. The test does not.