YCharts folds Zephyr SMA data into proposals; Advyzon ships AI suite

The August 2026 edition of Kitces' Nerd's Eye View reports that YCharts has acquired Zephyr, adding separately managed account data to its analysis and proposal tools. The same roundup reports that Advyzon has rolled out a suite of AI features inside its all-in-one platform. Both announcements are part of the column's monthly tracking of the advisor technology market.

On their faces, the two items are unrelated. One deepens the data behind client proposals; the other adds features to a platform advisors already log into daily. Together they make the same point: the next wave of advisor technology will ship inside the systems in use, not beside them.

The YCharts-Zephyr combination is about data depth. Separately managed accounts come with a manager, a fee, and a tax structure; the proposal is where that stack gets explained to a client. When the data lives in the tool that generates the document, the research step and the client-facing step become one workflow. That is a real efficiency, but it only holds if the data is complete and current.

The stakes are visible in the proposal itself. A client who sees a fee or tax assumption that contradicts the statement will ask why. When the data comes directly from the SMA manager through the proposal tool, the chance of that contradiction drops. But that only holds if the data pipeline is maintained and the acquisition is actually integrated, not just announced.

The Zephyr acquisition also signals that SMA data is being embedded at the point of sale. For years, advisors pulled research from one system and built proposals in another; the copy-and-paste risk was real. Moving the data into the proposal tool is a direct answer to that workflow break. The column's accompanying update to the AdvisorTech Solutions Map is a reminder that the advisor tool market is dense and shifting.

The broader pattern is familiar. Custodians have been layering on services, and TAMPs have been absorbing functions that were once separate products. Software that used to be sold as a point solution is becoming a module inside an all-in-one platform. For RIA principals, the question is no longer which tools have the best features; it is whether the platform you already pay for will add the feature you need before you need it.

Advyzon's AI suite shows how that question gets answered. Kitces frames the rollout as a sign that advisor AI is approaching mainstream adoption. The big incumbents have watched which features the early standalone tools made popular and are folding those picks into their existing platforms. That is a filtering process: standalone vendors become test kitchens, and the features that stick end up inside the platform that already has the daily login.

For an advisory firm, the distinction between buying new software and upgrading the software you already run is not cosmetic. A standalone tool means a procurement project: contract, security review, credentials, training, and a separate log-in. An in-platform feature is a training project with a much shorter list. The shorter list includes a new menu item and a walkthrough, not a third-party vendor relationship. But it still demands a judgment call — whether the built-in version matches the workflow the firm already runs.

The YCharts-Zephyr deal is a reminder that data depth is the battleground in proposal software. Model portfolios and SMA-style wrappers keep multiplying, and the proposal is where they get justified to clients. If the data behind the proposal is incomplete or stale, the proposal document carries that flaw. A firm that compares the built-in data against its own records before renewal will know whether the upgrade is real.

What firms should watch is whether these bundled features are actually used. Kitces notes that incumbents are picking features based on what's proven popular among early standalone tools. That suggests a filter at work, but also a risk that the built-in version is a compromise — a feature that works for the average firm but not for the one that depends on a specific edge case.

The renewal cycle is the natural checkpoint. A firm that runs a side-by-side test before renewal — building the same SMA proposal in the upgraded tool, running a real client file through the AI features — gets an answer based on its own work, not a vendor demo. The judgment will come down to data depth and feature completeness inside the platforms that already have the login.