Hi there,
Dashboards are starting to feel like homework. The fastest shift in SaaS right now is away from “click, filter, export” and toward something much simpler: ask a question and get an answer.
Conversational embedded analytics is a big part of that shift. Instead of shipping more reports, teams are building interfaces that let users talk to their data in plain language and get insights in seconds. In today’s lead story, we break down why NLQ is reshaping embedded analytics, what it takes to get it right in a multi-tenant SaaS environment, and how the best experiences move from one-off queries to real conversations.
📰 Upcoming in this issue
Talk to Your Data: How Conversational Interfaces Are Redefining Embedded Analytics
The Self-Service Shift: Your Users Expect It—and Yes, They’ll Pay for It
6 SaaS Trends in 2026
📈 Trending news
The AI Product Manager Playbook for SaaS Teams
2026 Dev Predictions for SaaS Product Leaders
SaaS Growth Isn’t Features: The Retention Leaks Product Leaders Miss
Talk to Your Data: How Conversational Interfaces Are Redefining Embedded Analytics
A lot of analytics friction comes from the same place: users have to translate a question into a dashboard workflow. Find the right report. Pick the right filters. Interpret the results. Export. Repeat.
Conversational interfaces flip that. With NLQ, users can ask questions like “What changed last week?” or “Where did engagement drop?” and get an answer instantly. That’s not just a nicer UI. It changes who can use analytics and how often it gets used.
Building a great conversational interface isn’t just about plugging in a language model. It requires:
Multi-tenant awareness: In SaaS, every tenant has different data. NLQ must respect those boundaries.
Domain tuning: Generic models don’t understand your product’s language. Fine-tuning is essential.
Explainability: Users need to trust the answers. That means showing how results were generated.
Fallbacks and guardrails: Not every question has a clean answer. Good design handles ambiguity gracefully.
The best experiences are also moving beyond single questions. They keep context, clarify intent, and guide users toward the next question. Over time, analytics starts to feel less like a reporting layer and more like a system that helps people understand what’s happening and what to do next.
Where this is heading: more context-aware conversations, multimodal inputs (text, voice, images), proactive insights that surface before someone asks, and tighter links between insights and actions inside the product.
Key Takeaways
🔍 The fastest path to adoption is fewer clicks. Plain-language questions reduce friction for everyday users.
🧠 Good NLQ is more than “text-to-SQL.” Context, clarification, and follow-ups make it feel useful.
🛡️ Trust determines usage. Clear explanations and guardrails matter as much as speed.
🚀 The next step is action. Expect analytics to increasingly trigger workflows, not just display insights.
The Self-Service Shift: Your Users Expect It—and Yes, They’ll Pay for It
Self-service analytics used to be a nice add-on, but it’s now the price of admission for B2B SaaS. End-user expectations have fundamentally changed: people want answers fast, in plain language, without filing tickets or waiting on customized dashboards built three sprints ago.
For SaaS teams, self-service is hard to build and risky to ship. Multi-tenant environments demand strict data isolation between tenants, fine-grained role-based access controls, and the need to securely benchmark or aggregate data across tenants without exposing sensitive information. That’s the tightrope!
Self-service typically shows up in two levels.
Beyond static dashboards: users can interact with embedded dashboards through filters, drill-downs, and drill-throughs to explore data dynamically.
True self-service: users customize dimensions, aggregations, visualization types on the fly, build charts from scratch using drag-and-drop tools, and share insights—no developer required.
AI as a force multiplier
The most transformative layer is AI-powered self-service. With NL prompts, users can ask questions like “Show me monthly trends by region” and get an auto-generated visualization, with an explanation of how the insight was derived.
AI also analyzes existing charts, surfaces anomalies, explains trends, and suggests follow-up questions. Analytics becomes guided exploration, not passive reporting.
And it pays.
Most SaaS companies monetize via value-based or usage-based subscription tiers—turning analytics from a cost center into revenue expansion.
Value-based pricing: basic tiers include pre-built dashboards and historical reporting; higher tiers unlock real-time data, more data sources, advanced self-service; and top tiers often include unlimited user-created content + AI-powered analysis.
Volume-based pricing: self-service access is limited by usage (# of reports, dashboards, users), with enterprise tiers removing limits and adding automation, forecasting, anomaly detection, and even AI-driven insights as add-ons.
Data sits at the heart of most SaaS products, yet it’s often underutilized. Self-service and AI adds tremendous value for your users, and positions your company to win.
Takeaways
🚀 Treat self-service analytics as core product functionality, not an add-on.
🎯 Plan AI integration intentionally, focusing on real user outcomes rather than novelty.
🛡️ Choose platforms designed for multi-tenant SaaS to avoid security and scalability pitfalls.
💰 Leverage analytics for monetization, using tiered or volume-based packaging.
6 SaaS Trends in 2026
The hows change every year, but the main drivers remain the same: Software-as-a-Service (SaaS) has to become smarter, faster, easier, and cheaper, and 2026 can play a noticeable part.
Market Growth & Macro Context
SaaS is not slowing down — demand remains strong. However, the growth becomes efficiency-driven rather than adoption-driven. Delivering ROI will become an important factor for adoption.
AI-Driven Evolution: “SaaS becomes smarter”
During the next decade, AI — especially generative AI, machine learning, and autonomous “agentic” systems — will support a significant shift from SaaS tools being merely cloud-based utilities to Human-in-the-loop autonomy, where AI executes, but humans approve.
Beyond "Basic" Analytics
Expect to see more SaaS products that don’t just visualize data and provide insights— but analytics that trigger workflows, and take action in an embedded fashion. Analytics stops being a reporting layer and becomes a decision engine inside SaaS products.
Shift Toward Vertical, Specialized, and Niche SaaS
The era of “one-size-fits-all” horizontal SaaS (e.g., generic CRM, general project management tools) is gradually yielding to more specialized, vertical-oriented, and domain-specific services — often deeply tailored to a given industry or solution shipped with:
Regulatory-aware data models
Pre-built metrics aligned to industry KPIs
Industry-specific benchmarks (peer comparisons)
Pricing & Business Models: More Flexibility, but Pressure
As SaaS evolves, we will see the rise of “outcomes-based” pricing — companies paying for results/ROI rather than just access to features. SaaS products will experience greater pressure to justify ROI. SaaS consumers may demand variable costs based on what their users need most. SaaS vendors will increasingly evolve their revenue models to accommodate AI workloads. Under the analytics and AI cost pressure, SaaS applications will provide tiered intelligence (not all users get the same AI depth), and also internal analytics to monitor ROI per tenant.
Economics, Competition, and Consolidation
Growth notwithstanding, there are several pitfalls and headwinds the SaaS market will face by 2026:
Margin compression for AI-first SaaS: AI workloads make SaaS operations more expensive, eroding the typical margins of classic SaaS.
Saturation and consolidation in horizontal SaaS: For many generic tools, growth slows as markets saturate. New user acquisition is becoming harder, especially as increasingly savvy buyers demand ROI.
Increasing demands on compliance, data security, and customizability — as SaaS enters more regulated verticals (healthcare, finance, etc.) — will put pressure to meet regulatory and privacy standards.
Why It Matters
SaaS companies that win are those that continuously add value. Right now, that means shrinking the distance between a user’s question and a trustworthy answer—inside the product, in seconds, without tickets, exports, or waiting on someone else. Embedded analytics and AI aren’t just “new features” in that context—they’re how products keep getting better after onboarding: faster time-to-answer, clearer decisions, and experiences that feel more helpful every week.
See you in the next edition,
Arman Eshraghi
The Embedded Intelligence Brief
Connect on LinkedIn