On this page
Every sales conference session, every vendor pitch, and a fair share of LinkedIn posts over the past two years have made some version of the claim that AI is transforming sales. Most of these claims are either significantly overstated or so vague they’re impossible to actually evaluate. The real picture is more specific and, in some ways, more interesting: AI has produced genuine, measurable change in certain parts of the sales cycle, left other parts almost untouched, and created a few new problems that didn’t exist before.
Where the Change Is Real: Research and Preparation
The clearest, least disputed impact of AI in sales sits in pre-call and pre-outreach research. What used to take a rep or SDR twenty to thirty minutes per account, digging through a company’s website, recent news, LinkedIn activity, and tech stack signals, can now be compiled into a usable summary in a fraction of that time. This is a genuine efficiency gain, not a hype claim, and it shows up clearly in how much more targeted outbound has become across teams that have adopted these tools properly.
The risk here isn’t that the tool doesn’t work, it’s that the efficiency gain gets treated as a substitute for judgement rather than an input to it. A rep who reads an AI-generated account summary and treats it as complete understanding, without applying their own read on what actually matters for that specific buyer, ends up with research that’s fast but shallow. The reps getting the most value are using AI research to compress the time spent gathering information, then spending the time saved on actually thinking about strategy and approach, rather than skipping that thinking step entirely.
Where the Change Is Real: Outbound Volume and Personalisation at Scale
AI has meaningfully changed what’s possible in outbound sequencing, particularly around producing individually relevant-sounding messaging at a volume that would have required significant manual effort previously. This is a double-edged development. It’s raised the floor on outbound quality for teams using it well, and it’s also flooded buyer inboxes with a much higher volume of AI-generated outreach overall, which has made genuinely differentiated, well-researched messaging more valuable precisely because there’s more mediocre AI-assisted noise to stand out against.
Buyers, including ANZ buyers who are generally quick to spot generic personalisation, have become noticeably better at identifying AI-generated outreach that wasn’t properly reviewed or refined by a human before sending. Teams that use AI to draft a first pass and then genuinely edit for specificity and voice are seeing this tooling work well. Teams that send AI output largely unedited are seeing reply rates that have gotten worse, not better, as buyers’ pattern recognition for generic AI messaging has sharpened.
Where AI Has Barely Moved the Needle: Complex, Multi-Stakeholder Deals
Despite a lot of confident claims about AI transforming enterprise sales, the actual mechanics of managing a complex, multi-stakeholder deal, reading a room during a tense internal stakeholder meeting, sensing when a champion’s enthusiasm doesn’t match their actual internal influence, navigating a procurement negotiation with genuine tactical judgement, remain almost entirely dependent on human skill. AI tools can support this work with better data and preparation, but the actual judgement calls that decide whether a complex enterprise deal closes are still squarely a human capability, and there’s little evidence that’s changing quickly.
This matters specifically for ANZ, where enterprise and mid-market deals tend to be more committee-driven and relationship-dependent than faster-moving transactional SaaS motions common in some other markets. Teams that have over-invested in AI tooling on the assumption it would meaningfully speed up or de-risk this stage of the sales process are generally finding the return far lower than what they saw from AI investment in top-of-funnel activity.
Forecasting and Deal Risk Scoring: Promising, Still Immature
AI-powered deal scoring and forecasting tools, which analyse email sentiment, meeting frequency, and engagement patterns to flag deal risk, have improved meaningfully but remain genuinely mixed in practice. They’re useful as an early warning signal, flagging deals with declining engagement or unusual communication patterns that might otherwise go unnoticed until a manager’s next scheduled deal review. They’re considerably less reliable as a standalone predictor of close probability, particularly for the more complex, relationship and committee-driven deals common in ANZ enterprise sales, where the signals these tools rely on, largely email and calendar activity, capture only a partial picture of what’s actually happening inside a buying committee.
The teams getting genuine value from these tools are using them as one input into human-led deal review, a prompt to ask better questions in a deal review conversation, rather than as an automated substitute for that conversation. Teams treating AI risk scores as a replacement for genuine manager judgement in deal review are generally getting worse forecast accuracy, not better, because the tool’s blind spots become the team’s blind spots.\
What This Means for How Teams Should Actually Invest
The practical takeaway isn’t that AI investment in sales is overhyped across the board, parts of it are producing real, measurable efficiency gains that any team would be leaving value on the table by ignoring. The more useful frame is being specific about where AI is actually earning its keep, research compression, outbound efficiency, early risk signalling, versus where it’s being deployed on the assumption of impact that the evidence doesn’t yet support, particularly around complex deal management and stakeholder navigation.
Teams making the most of AI right now tend to share a similar pattern: they’re aggressive about adopting it in the parts of the sales cycle where the efficiency gain is proven and significant, and they’re deliberately conservative about assuming it replaces judgement in the parts of the cycle, discovery depth, objection handling, multi-stakeholder navigation, negotiation, that still depend on a skilled human reading a genuinely complex, ambiguous situation in real time. Getting that balance right, rather than either dismissing AI’s real value or overestimating what it currently does, is what’s actually separating teams seeing genuine productivity gains from teams that have added a lot of new tooling without much to show for it.
Figuring out where you fit on this map? Browse open GTM roles to see how these titles show up in real job ads, or check the Career Advice hub for more on breaking into GTM.
