The Next Salesforce Moment Is Happening Right Now
How AI-native sales software could challenge today's SaaS giants, and why the ability to adapt to individual users might become the next major competitive advantage.
Patrick Trümpi9 min read
- Sales enablement
- Taskbase basics
The Last Salesforce Moment
In 1993, Siebel Systems was founded and became one of the most important CRM providers in the world. Six years later, Salesforce entered the market.
Today, Salesforce is one of the world's largest software companies, while Siebel was acquired by Oracle in 2006. Considering Siebel's six-year head start, that is quite remarkable.
What happened?
The internet made it possible to deliver enterprise software in a completely different way. Until then, companies installed CRM systems on their own servers, spent substantial amounts on implementation, and often struggled to upgrade them.
Salesforce introduced a cloud-based model where customers could access their CRM through a browser and receive improvements without installing new versions themselves.
I believe one advantage was particularly important: Salesforce could develop and distribute new functionality much faster than traditional software providers.
A Siebel customer might wait months or years before upgrading to a new version. Salesforce could roll out improvements centrally, eventually establishing three major releases per year.
Combined with a subscription model that lowered upfront costs, this helped Salesforce reach smaller companies and grow alongside them.
Of course, distribution, marketing, and execution also played important roles. But I want to focus on the speed at which software could improve and reach customers.
Because I believe we are approaching another shift of similar importance.
And this time, some of the companies that benefited most from the previous revolution could find themselves on the other side of it.
AI Is Changing the Economics of Building Software
During 2025 and 2026, AI agents became considerably more capable.
We have moved from systems that primarily generate text and summarise information towards agents that can execute workflows, update CRM records, analyse data, and interact with multiple applications.
At the same time, vibe coding has made it remarkably easy to create software interfaces through natural-language instructions.
Building reliable enterprise software still requires security, permissions, integrations, and considerable engineering effort. Nevertheless, the time and cost required to develop certain types of functionality have changed dramatically.
I believe the implications go much further than making developers more productive.
We are moving towards software that can adapt to individual users while they are using it.
And that could fundamentally change how enterprise software companies compete.
Why We Couldn't Compete With Gong in 2021
In 2021, while working at Unique, we explored building a product that would compete with Gong.
We developed recording capabilities, conversation analysis, and interfaces where salespeople could review calls and understand their opportunities. We also wanted to build deal risk assessments, coaching functionality, and better pipeline visibility.
The problem was that Gong had a substantial head start, considerably more engineers, and much greater resources.
Every new use case required us to build another interface. Pipeline management needed a pipeline view. Coaching required scorecards. Deal qualification required dashboards.
Gong was doing exactly the same thing, only much faster.
We eventually moved away from that market and focused on financial services.
But something else stayed with me.
Many users were primarily interested in recording conversations and watching them afterwards. Despite all the functionality we developed, they were using only a relatively small part of it.
That made me question whether continuously adding more features was actually the best way to help salespeople.
Today, I think this question is more relevant than ever.
The Problem With Traditional Sales Software
Look at modern revenue platforms such as Gong.
They offer conversation intelligence, forecasting, deal management, coaching, reporting, and numerous other capabilities. For large organisations, this functionality can be extremely valuable.
But how much of it does the average salesperson actually use?
I don't have Gong's internal usage numbers, but based on my experience building similar software, I suspect the gap between available functionality and regularly used functionality can be substantial.
An account executive already spends their day in a CRM, emails, meetings, and presentations. Introducing another platform with dozens of dashboards requires them to understand what is available and remember where to find it.
And different users have very different needs.
A CRO might want a detailed forecasting dashboard. An account executive might simply want to know which opportunities require attention this week.
Traditionally, software companies have tried to satisfy these requirements by building more interfaces.
AI sales agents create another possibility.
From Standardised Interfaces to Personalised AI Sales Software
At Taskbase, we have started thinking about software functionality as apps.
An app combines data sources, business logic, and a way of presenting information. A forecasting app might aggregate predictions across a sales organisation. A deal risk app might evaluate opportunities against the company's qualification methodology.
These apps can still have traditional interfaces.
But imagine an account executive simply asking an AI agent which deals deserve attention this week.
The agent knows the opportunities, upcoming meetings, communication history, and qualification criteria. It can generate a view containing precisely the information the salesperson needs.
If they want to see only deals above €50,000, opportunities without a next meeting, or deals approaching a critical event, they can simply ask.
The same principle applies to meeting preparation.
Personally, I prefer detailed meeting briefings. I want to understand the opportunity history, stakeholders, challenges, and objectives.
Other salespeople prefer a few bullet points.
Traditionally, software companies would build different templates and configuration options. With an AI agent, users can explain their preferences, and the software can adapt.
We apply a similar approach at Taskbase when generating follow-up emails. By analysing previous emails, the AI can better understand an individual salesperson's writing style while following the company's communication guidelines.
Over time, the agent can learn from corrections and feedback.
This creates an interesting possibility: software that improves the individual user's experience without requiring a new feature release for everyone.
The Sales Playbook Becomes the Foundation
There is an important condition for making this work.
If every salesperson interacts with software differently, how does the company maintain a consistent sales process?
I believe the answer lies in the sales playbook.
And I mean something considerably more comprehensive than MEDDICC or SPICED.
A proper sales playbook defines the problems a company solves, its positioning, sales processes, discovery approach, messaging, objection handling, competitive intelligence, and the skills required for different roles.
Historically, maintaining this knowledge was difficult. Playbooks became outdated, and sales leaders struggled to keep hundreds of pages of information relevant.
AI changes what is possible.
An AI sales agent can retrieve the appropriate guidance for a specific situation, while analysing actual conversations to identify new objections, successful approaches, and potential improvements.
Imagine several salespeople encountering the same objection. The system identifies how top performers respond and suggests an update to the playbook.
Once the sales leader approves it, the guidance becomes available to everyone.
The organisation maintains consistent standards while individual employees receive personalised support.
Why This Matters for AI Sales Coaching
I believe the consequences for sales coaching are particularly significant.
Traditional sales software can identify weaknesses in discovery, qualification, or objection handling. But salespeople still need to find that feedback, interpret it, and decide how to improve.
In my experience, even excellent salespeople rarely spend their working day exploring coaching dashboards.
A proactive AI sales coach can approach this differently.
Imagine an account executive preparing for an important discovery call. The AI coach knows that this person has recently struggled to quantify the financial impact of customer problems.
Before the meeting, it suggests a few relevant questions based on the company's sales playbook and the opportunity context.
Afterwards, it analyses the conversation and provides specific feedback.
Perhaps the salesperson explored the operational problem well but never established its economic consequences.
The coach can identify that weakness and help the salesperson improve before the next conversation.
This is the direction we are pursuing at Taskbase. Coaching becomes part of everyday work, connected to actual customer interactions and the skills each salesperson needs to develop.
The same approach can help AI agents automate administrative work, such as preparing CRM updates or suggesting next steps.
Of course, the system must learn when to intervene. An AI coach that constantly interrupts people with irrelevant suggestions will quickly become annoying.
But when the timing and context are right, the potential impact on adoption and performance is substantial.
Could AI-Native Platforms Challenge Today's Software Giants?
This brings me back to Salesforce and Siebel.
Established platforms such as Gong and Salesforce have enormous advantages. They have extensive functionality, valuable data, deep integrations, and large customer bases.
They also have substantial amounts of existing software to maintain.
Every interface and workflow introduces complexity. Changes need to work across thousands of customers and different configurations.
AI-native platforms can start with a different architecture, combining shared data, reusable applications, and agents that adapt the user experience.
A company with 20 salespeople might need basic forecasting, conversation analysis, and coaching. A company with 2,000 salespeople might require considerably more sophisticated processes.
Both could potentially use the same underlying platform, with functionality adapted to their needs.
That is particularly interesting for smaller and mid-sized companies that currently purchase software containing far more functionality than they actually use.
Of course, established vendors are investing heavily in AI themselves. They could make this transition successfully.
The question is how quickly they can adapt their existing products and operating models.
The Next Competitive Advantage Is the Ability to Adapt
When Salesforce entered the market, the cloud allowed it to improve and distribute software faster than many established competitors.
AI agents could create a similar advantage, this time at the level of the individual user.
A CRO can have the forecasting dashboard they need. An account executive can ask which deals deserve attention. An SDR can receive coaching before a cold call.
All three can work with the same company data and sales playbook, while interacting with software in different ways.
Historically, having more engineers and more years of development often meant having more functionality. I experienced how difficult that advantage was to overcome when we tried to compete with Gong in 2021.
Today, I am much less convinced that the number of interfaces a company has developed over the past decade will remain such a powerful competitive advantage.
The quality of the underlying data, the reliability of AI agents, their understanding of company-specific processes, and the speed at which software adapts to users may become considerably more important.
Building reliable enterprise software will remain difficult. But the economics of creating and adapting functionality are changing.
And I believe the companies that embrace this change fastest could have an enormous advantage over the next decade.
I think we are witnessing the beginning of another Salesforce moment.
At Taskbase, we are building AI-native sales coaching that connects a company's sales playbook with real customer interactions, helping salespeople improve their skills through personalised guidance in their daily work.