When “Good Enough” Stops Being Good Enough
ThoughtSpot built its reputation on search-driven analytics and a bold promise: give business users access to data without needing a data team standing behind them. That promise still holds weight. But a younger competitor, Omni, is quietly winning over the engineering teams who actually build the products ThoughtSpot’s customers use – and it’s doing it with an API-first architecture that ThoughtSpot’s embedded offering has never quite matched.
The Embedded Analytics Problem Nobody Talks About
Embedded analytics is not a niche market. It sits at the center of how modern SaaS products deliver value – dashboards inside CRMs, usage reports inside project management tools, financial summaries inside accounting platforms. The expectation from product teams has shifted: analytics should feel like it belongs inside the product, not bolted on from the outside. That distinction, between “integrated” and “embedded,” is where ThoughtSpot’s architecture starts to show friction.
ThoughtSpot Everywhere, the company’s embedded offering, gives developers a way to drop analytics components into external products using iframes and JavaScript embed codes. It works. But working and winning are different things. The iframe model carries a ceiling – styling constraints, limited event handling, and a control surface that stops at the edge of ThoughtSpot’s own UI layer. Engineering teams building opinionated product experiences find themselves fighting the tool rather than building with it.
Omni’s approach treats the API as the primary interface rather than a secondary export path. Developers query Omni’s semantic layer directly, pull structured data into their own rendering layer, and build whatever interface makes sense for their product. There is no iframe to work around. The UI is whatever the engineering team builds. That is a fundamentally different relationship between a vendor and the developers consuming its product.
The practical consequence is that Omni fits cleanly into modern frontend stacks – React, Next.js, Vue – without requiring developers to manage a visual component they cannot fully control. ThoughtSpot’s embedded SDK has improved over successive releases, but the core model still anchors the experience to ThoughtSpot’s own rendering engine. For product teams who care deeply about interaction design, that anchor matters.
Why the API Layer Is Winning the Developer Audience
Developer-led adoption has rewritten how analytics vendors compete. A decade ago, the buying decision lived with a VP of Business Intelligence who evaluated feature checklists and sat through demo calls. Today, a senior engineer evaluating tools for a new product build will spin up a trial, hit the API documentation, and make a judgment call within a few hours. If the API is thin or awkward, that engineer moves on. Omni understood this dynamic early.
Omni’s semantic layer – the governed, reusable definitions of metrics and dimensions that sit between raw data and the querying surface – is accessible through its API without requiring a user to be inside Omni’s interface at all. That means a company can define “monthly recurring revenue” once inside Omni, then reference that definition from any downstream product or tool that calls the API. ThoughtSpot has a semantic layer, too, but exposing it through a programmatic API in a way that developers can cleanly consume without touching the UI has been a slower, more constrained evolution.
The pattern playing out resembles what happened when Stripe entered the payments market against incumbents who offered hosted payment pages. Stripe’s API gave developers control over every pixel and interaction. Incumbents offered hosted flows that were functional but opaque. Developers voted with their integration decisions. Omni is making a similar bet: that control over the rendering layer is not a luxury for product-focused engineering teams, it is the baseline requirement.
There is also a pricing tension worth examining. ThoughtSpot’s embedded offering scales on consumption and user seats in ways that can become expensive as a customer’s end-user base grows. For a SaaS company embedding analytics for thousands of customers, seat-based pricing is a liability that compounds with growth. Omni’s pricing model has drawn attention from founders and product leaders specifically because it does not punish growth in the same way. When the analytics layer’s cost scales faster than the revenue it helps generate, the tool becomes a business problem, not just a technical one.
It would be an overstatement to call ThoughtSpot’s embedded product broken. Large enterprises with existing ThoughtSpot deployments are not canceling contracts. But the new business – the Series B SaaS company building its first analytics product, the fintech startup adding reporting to its dashboard – is increasingly starting with Omni rather than evaluating ThoughtSpot as a default. That shift in where new projects begin is the kind of quiet erosion that takes years to show up in market share numbers but changes competitive dynamics well before the data does.
ThoughtSpot’s Remaining Strongholds
ThoughtSpot is not without advantages that Omni has not yet matched. Its natural language querying – the core product that made the company – still outperforms most competitors in environments where non-technical business users need self-service access to complex data. For internal analytics deployments, executive dashboards, and large enterprise BI rollouts where the end user is an employee rather than a customer, ThoughtSpot’s experience remains strong. The search-driven interface has real advocates inside companies that have standardized on it.
The harder question is whether those use cases – internal analytics for business users – represent where the market is growing, or where it has already matured. The product analytics and embedded reporting markets are expanding faster than traditional internal BI. If ThoughtSpot’s strongest ground is also its most crowded and slowest-growing ground, Omni does not need to beat it everywhere. It only needs to own the part that matters to developers building new products, which is exactly the part it is currently winning.
