Melty, the AI Coach designed to reduce cognitive load for SaaS users

SaaS products keep getting richer, but long-term customer adoption remains a challenge. This article explores how in-product, contextual guidance can help software vendors improve activation, usage, and retention, without adding friction to the product experience.

Every new SaaS product comes with the same promise: more performance, more efficiency, more automation for customers. On paper, everything looks right. Products are richer in features, better integrated, and able to support an ever-growing range of use cases.

Yet from the user’s perspective, reality is often more nuanced. As products evolve, users are expected to absorb more functionality, navigate increasingly complex workflows, and adapt to frequent changes. The result is an invisible but very real friction that directly impacts activation, adoption, and ultimately retention.

Today, most of our working days are spent inside software. CRMs, ERPs, HR tools, marketing platforms, financial systems, and internal tools shape how work gets done and decisions are made. For software vendors, however, one question remains largely under-addressed: how do you help customers use your product well over time, based on their real-world usage, without adding friction to the product experience?

This is exactly the question that drives us at MeltingSpot (and the one that genuinely motivates us).

In most SaaS companies, user enablement efforts are concentrated at the very beginning of the customer journey. Onboarding sessions, product walkthroughs, help centers, sometimes a few product tours. Then the product evolves, new features are released, use cases expand… and supportive guidance gradually fades away.

In our conversations with SaaS users, expectations are remarkably consistent. They are not asking for more content, more training sessions, or another tool to consult. What they really want is help at the right moment, directly inside the product they are using.

They want to understand what they are doing, why they are doing it, and how to use a specific feature correctly, without being pulled out of their workflow. In practice, reading long, generic documentation is not an option. Static product tours quickly lose relevance, and LMS-hosted training is often accessed too late, or not at all.

What many users are looking for, very pragmatically, is simple, in-product guidance: something that can answer common questions, reinforce best practices, and suggest better ways to use the product when it actually makes sense.

Before going deeper, a demonstration is often the most effective way to make this concrete. The video below shows how Melty integrates directly into a SaaS product to guide and train users throughout their day-to-day usage.

Join the Melty beta by signing up here.

Many solutions already exist to address product adoption challenges from a vendor’s perspective. Each tackles part of the problem, but rarely the whole picture.

Traditional Digital Adoption Platforms still rely heavily on static logic: predefined paths, step‑by‑step instructions triggered independently of the user’s real context. Yet within the same SaaS product, usage can vary dramatically depending on user profiles, maturity levels, and customer objectives.

Support chatbots, including AI‑powered ones, remain mostly reactive. They wait for explicit questions and only cover a subset of situations where users might need guidance. As for LMS platforms, they play a useful role in structuring training, but by design, they sit outside the product and outside the flow of work.

At MeltingSpot, we have spent the past eighteen months working on a very concrete challenge for software vendors: how to improve SaaS product adoption without adding friction to the user experience, and how to help customers build skills without multiplying heavy or disconnected training initiatives.

We first built a digital adoption and user training platform, with in‑app content, learning paths, micro‑learnings, and tutorials embedded directly into software. These foundations were necessary, but not sufficient. One key element was missing: a single coordination point, fully dedicated to the end user.

That insight led to Melty, an AI Coach embedded directly inside SaaS products. It is neither a generic chatbot nor a simple overlay, but a coach designed to observe real product usage and intervene in a targeted way.

Melty analyzes behaviors, identifies frictions or misunderstandings, suggests relevant actions or content, and engages in conversation when it truly adds value. The goal is not to interrupt users, but to support them proactively, contextually, and with restraint.

In practice, SaaS customers often find themselves torn between initial onboarding sessions, a dense knowledge base, and interactions with support teams. Over time, the issue is rarely motivation or engagement, it is the gap between when help is available and when it is actually useful.

This is where context, repetition, and continuity make all the difference.

Melty is currently in beta. We are gradually opening access to SaaS teams who want to test this new approach to product adoption and help shape its evolution. If this topic resonates with you, you can sign up to join the beta.

Integrating an AI Coach directly into a SaaS product shifts several key dynamics. Users gain autonomy, best practices are reinforced in context, usage quality improves, and first‑line support requests decrease. Adoption becomes continuous rather than event‑based, and far more sustainable over time.

An AI Coach is not meant to replace Customer Success or Product teams. It acts as a relay, handling recurring explanations, reminders, and fundamentals, allowing teams to focus on higher‑value challenges such as deep usage understanding, advanced activation, and retention.

Melty is still a work in progress, and that is a deliberate choice. We want to build it alongside software vendors facing these challenges every day. A new way of supporting SaaS users is emerging: more integrated, more contextual, and more aligned with real usage. The next chapter is being written now.