The latest 2Checkout Live Panel brought together practical ideas for software and SaaS teams from people who are working with AI every day.
The live panel discussion focused on what companies can do right now to get real results with AI, without ending up with a messy stack of tools and a lot of half-finished experiments.
Featuring insights from Frank Sondors at Salesforge, Sara Maldon at Make.com, and Mike Korba at User.com, the session shared clear, grounded advice for teams trying to move faster and stay focused.
Key highlights include:
- Why it’s easier than ever to build software today, and why that makes quality matter;
- How AI can boost profits by helping teams get more done with fewer hires;
- Where AI is already paying off quickly, and where it still struggles;
- A simple way to decide what to automate, what to use AI to support, and what should stay human-led;
- How to try new tools without collecting subscriptions you never use;
- Why adoption stalls most AI projects, and how clear training plus support for middle managers keeps change moving.
What follows is a recap of the panel’s strongest lessons, with examples you can use.
The AI Trends That Will Move the Business in 2026
Margins are the Quiet Win
One of the clearest themes from the panel was that AI is already changing the shape of SaaS businesses, even when revenue doesn’t look wildly different quarter to quarter.
Frank Sondors puts it simply: the quiet win is in margins.
In most software businesses, people are still the biggest cost. When teams can get the same work done with fewer hires, profitability shifts fast.
AI doesn’t just help you do more, but rather changes how many people your business needs to hire to hit the same targets, and that affects how a company grows, what it can afford, and how it funds that growth.
Frank also frames AI progress in “waves”:
- First came text, where you put a prompt into a model like ChatGPT and get usable text back.
- Then came voice, with tools like ElevenLabs as an example of how quickly that space is moving.
- And the newest wave, Frank argues, is video: video models can now create convincing deepfakes, and the result can look realistic enough that most people wouldn’t spot the difference.
Competition is Multiplying
Another point Frank makes is that building software today is cheaper and easier than it was even a year or two ago.
Both Frank and Sara Maldon point to the same result: more new products, more niche tools, and more companies are trying to win a small slice of your market.
That has a real knock-on effect on retention. It’s not only about losing net-new deals. It’s also about customers having more options, and switching feeling easier when a small tool solves one specific problem really well.
Quality Becomes the Moat
During the discussion, Sara makes a solid point: When it comes to SaaS, it’s no longer enough to solve a problem. Your business has to solve it well, and in a way that customers feel great to use.
In a crowded market, the difference is often the basics, but done better:
- A smoother product experience;
- Clearer workflows;
- More consistent results;
- Less friction for the customer.
She also tied this to personalization. With better access to context and data, SaaS companies can deliver experiences that feel more tailored and less generic, which becomes a lever for retention and expansion.
Pricing is Changing Under SaaS
Mike Korba brings up an important trend: pricing models are being forced to evolve.
For years, seat-based pricing was the default in many SaaS categories. Mike argues that now more customers are accepting usage-based pricing, and companies are learning why that matters.
If your cost base rises with consumption, “unlimited” becomes a margin risk. The model needs to reflect cost reality, especially when AI-driven usage and other behind-the-scenes workloads can scale unpredictably.
Retention Gets Smarter
The speakers also pointed to a clear retention shift: AI can help teams spot churn risk earlier and reach out in a way that feels personal, even when they’re doing it across hundreds or thousands of customers.
Some benefits mentioned are::
- Predicting churn risk sooner;
- Segmenting customers more intelligently;
- Reaching out in a way that feels more personal, even across a large customer base.
Done well, it’s less about increasing automation and more about scaling the level of attention that was once practical only for a small set of strategic accounts.
Where AI Is Paying Off Now – and Why
Customer Support is the Quickest ROI
Mike mentions that customer support is often the quickest place to see ROI from AI because it’s accessible and easy to roll out.
With tools that can handle repeatable questions, teams can see impact early, sometimes within the first day, week, or month, which is why so many SaaS companies start there.
Sales and Revenue Ops: High-Volume Work That Compounds
The topic of sales came up as a main priority throughout the discussion panel.
Frank makes it a point that every company needs a pipeline. He shares two examples his team is rolling out:
- agentic demos, because they run hundreds of demos each month,
- and agentic CSM to support thousands of long-tail accounts that don’t typically get dedicated coverage.
Agentic Demos and Agentic CSM for Scale
The thread connecting both examples is pretty straightforward: customers want a more personal experience, but most teams can’t afford to staff that level of attention for every account.
If AI can take on the first pass, handle the routine questions, and keep things moving, then companies can give more customers timely support without turning the team into a hiring plan.
Frank describes this as a sequence he tries to follow. In his own words:
- Start with an agent tool if one already exists;
- If it doesn’t, build a simple workflow;
- If building it is too much, bring in a freelancer or an agency;
- Only hire once it’s clear the work needs long-term ownership.
HR and Employee Experience are Underrated
Sara points to an under-discussed area beyond sales and support: HR and people operations.
She argues that many workflows remain unnecessarily manual, and that better systems can materially improve how employees experience the organization.
Examples include:
- onboarding, so new hires get the right information at the right time without chasing people for answers;
- day-to-day people ops tasks that eat up HR time but don’t add much value;
- better visibility into how the workforce is changing, so talent movement is less reactive and more predictable.
Automate, Augment, or Keep it Human
One of the most interesting parts of the panel is how practical the “what do we do with AI?” question becomes once businesses stop treating it like a single decision.
As Mike Korba puts it, the real choice isn’t AI vs. no AI, but rather where you automate fully, where you use AI as support, and where you keep people in control because the context is too important or too unpredictable.
He suggests looking at any workflow through a simple lens to make this decision:
- Volume: how often does this happen?
- Risk: what’s the downside if the output is wrong?
- Complexity: is the process straightforward, or does it change case by case?
- Judgment: does it require human decision-making and context?
Humans for Ambiguity, Agents for Routine Work
Frank backs up Mike’s point with a pattern you see in real-world rollouts: AI works best when the context is clear. But when the situation is open-ended, humans still outperform.
He gives a few easy ways to apply this to any SaaS business:
- Early conversations, where you don’t know what you’re walking into, tend to stay human-led;
- Later-stage moments, where the routine is known and the goal is clear, are a better fit for AI agents.
A good example of where voice agents can work well is outbound calls to trial users who signed up but didn’t convert.
In this case:
- the audience is known,
- the situation is consistent,
- and the goal is clear: understand what got in the way, capture that feedback at scale, and lastly, route qualified prospects back into the funnel when it makes sense.
Frank also warns against applying AI in situations where judgment is the main requirement.
Support is a clear example. AI can be effective for common questions with answers that already exist in documentation. But when tickets are driven by bugs, unusual edge cases, or something truly broken, it still requires real humans to diagnose the problem and resolve it.
Agents Work Best With A Solid Foundation
Sara talks about pushing back on a trend she sees often: teams asking for agents when what they really need is better automation.
She argues that a lot of business problems don’t actually require an agent making decisions. They require a cleaner process.
Her view is that agents usually work best when they sit on top of strong foundations:
- clear workflows;
- good data;
- basic automations already doing the predictable work.
It’s important to note that if a business decides to skip these foundations and jump straight to an agent, it can end up with something expensive and unreliable trying to solve a problem that wasn’t well defined in the first place.
Choosing the Right AI Tools
Business-Critical vs. Optional
Frank mentions that tool overload isn’t happening because teams aren’t smart buyers, but rather because there are too many options, too little time, and a steady stream of “must-try” tools on the market.
He shares a simple filter that keeps decisions around SaaS grounded. As he puts it – treat tools like painkillers or vitamins.
Painkillers solve problems businesses can feel right now. Vitamins, on the other hand, sound good, but they rarely change outcomes.
And if a tool can’t connect back to something concrete, it’s likely not worth adding.
- revenue and pipeline;
- retention and customer experience;
- margin improvement through time and headcount savings.
The Build vs. Buy Playbook
Frank also describes an order of operations for solving problems that keeps teams fast and lean:
- buy a tool off-the-shelf if it’s ready and proven;
- if it isn’t, build the workflow internally;
- if the tool is too specialized, outsource to an agency or freelancer;
- hire only when it’s clear you need long-term ownership in-house.
Clean Up Your Stack Regularly
Both Frank and Mike Korba stress a step many companies skip: you need a way to remove tools, not just add them. Without that, subscriptions accumulate, and over time it becomes harder to tell what’s truly necessary.
A simple approach to follow is this:
- assign someone ownership of subscriptions and renewals;
- review tool usage regularly;
- if a tool isn’t being used, cancel it quickly instead of letting it linger.
Experiment Without Creating Chaos
Mike suggests separating the “must-have” systems from the test-and-learn tools. Your core platforms need more discipline and alignment, but teams still need room to experiment, ideally with a small, clear budget for trials.
Sara adds a practical way to keep experimentation from becoming sprawl: if multiple tools solve the same job, run a short competition and pick a winner. One tool becomes the standard. The rest get dropped.
Why AI Projects Stall, and How to Keep Them Moving
Culture and Fear
Most of the time, what slows AI down isn’t the software, but the people’s side of change. Frank and Mike both point to the same reality: even in companies where AI is a daily topic, plenty of employees still aren’t using the basic tools in a steady, confident way.
Some simply don’t know where to begin. Others hold back because they’re uneasy about what AI could mean for their job.
That’s why adoption can’t be treated like a quick launch and a link to a new tool. It takes hands-on support, simple training, and clear messaging that the goal is to help people do better work.
Middle Managers Carry the Load
Sara points out that the real pressure sits in the middle. Leaders usually want progress, while teams feel uncertainty.
Middle managers have to hold both at once.
They’re the ones translating strategy into daily habits, answering hard questions, and keeping people steady while workflows change. When that layer isn’t supported, AI efforts stall, even if leadership is fully bought in.
Next Steps: Turning AI Noise Into Real Progress
If there’s one clear message from this panel, it’s that AI progress doesn’t come from chasing every new release.
Instead, it comes from making a few smart choices and sticking with them.
Focus on the areas where AI is already proving its value, keep humans in charge where judgment and trust matter, and be disciplined about the tools you bring into the business.
This recap covers the main takeaways, but the full panel includes more context, examples, and nuance from each speaker.
Watch the complete AI at Full Speed session from 2Checkout to hear the full discussion and turn the ideas into practical next steps for your team.
The post AI at Full Speed: How Software & SaaS Companies Accelerate Results in 2026 – Live Panel Wrap-Up appeared first on The 2Checkout Blog | Articles on eCommerce, Payments, CRO and more.
This articles is written by : Nermeen Nabil Khear Abdelmalak
All rights reserved to : USAGOLDMIES . www.usagoldmines.com
You can Enjoy surfing our website categories and read more content in many fields you may like .
Why USAGoldMines ?
USAGoldMines is a comprehensive website offering the latest in financial, crypto, and technical news. With specialized sections for each category, it provides readers with up-to-date market insights, investment trends, and technological advancements, making it a valuable resource for investors and enthusiasts in the fast-paced financial world.