From Rejection to Connection: How AI Is Changing Sales for Good [Podcast]

TL;DR

• AI doesn’t replace salespeople — it removes the administrative weight that stops them from doing their best work: building trust and understanding what buyers actually need.
• The biggest gains from AI in sales come from three areas: pre-call research, lead qualification, and follow-up consistency — not from replacing human judgment.
• McKinsey and Forrester research consistently points to productivity gains in the range of 20–40% when sales teams adopt AI-assisted workflows, though results vary significantly by implementation quality.
• For SMBs, the risk isn’t adopting AI too early — it’s adopting the wrong tools without a clear process to support them.
• The goal isn’t automation for its own sake. It’s giving salespeople more time in front of the right people, having better conversations.

---

What “AI in Sales” Actually Means for a Small Team

Sales has always been a contact sport. It runs on timing, trust, and the ability to understand what someone actually needs — often before they can articulate it themselves. Nothing about that changes with AI.

What does change is everything around those moments. The research before the call. The follow-up that slips through the cracks. The qualification call that takes 45 minutes and goes nowhere. The CRM entries that never get done. These are the places where AI earns its keep for a sales team.

For a founder-led business with five to twenty people in a sales or business development function, the math is simple: your team spends a meaningful portion of their week on work that isn’t actually selling. AI can address that gap. Not perfectly, not immediately, and not without effort — but meaningfully.

The framing that matters here is augmentation, not replacement. A recruitment firm’s account manager who used to spend two hours building a prospect brief before a client call can now do it in fifteen minutes, with better data. That’s not a job threat. That’s two hours back per day to spend on conversations that actually move deals forward.

---

The Three Highest-Value Applications in an SMB Sales Context

Not all AI use cases are equal. For smaller teams with limited resources, it pays to be selective. The following three applications consistently deliver the clearest return, regardless of industry.

1. Pre-call research and briefing

AI can pull together a structured brief on a prospect — recent company news, likely pain points based on firmographic data, relevant context about their industry — in a fraction of the time it takes manually. For a commercial real estate broker, a legal services firm, or an accounting practice, walking into a discovery call with that level of preparation changes the quality of the conversation immediately.

This isn’t about impressing prospects with how much you know. It’s about not wasting their time. Buyers notice when a salesperson is genuinely informed. They also notice when they’re not.

2. Lead qualification and scoring

Most SMB sales teams have a lead problem — not a shortage of leads, but a quality and prioritisation problem. They’re spending time on prospects who were never going to buy, while high-intent leads sit waiting too long for a response.

AI can analyse behavioural signals — email engagement patterns, page visits, form completions, prior interaction history — to surface which prospects are worth prioritising today. This isn’t magic. The model is only as good as the data going into it. But even a rough signal is more useful than gut feel alone, particularly when your team is stretched.

3. Follow-up consistency

This is where most SMB sales processes leak revenue. A prospect shows interest, there’s a good call, and then the follow-up either never happens or arrives three days later with a generic message. Life gets in the way. Deals die in the silence.

AI-assisted follow-up sequences — where the timing, content, and personalisation are automated but reviewed by a human before sending — solve this problem without making outreach feel robotic. The salesperson sets the tone and approves the message. The AI handles the logistics of when, what, and to whom.

---

What Good Implementation Looks Like

The gap between “we installed an AI tool” and “AI is actually improving our sales performance” is almost always a process gap, not a technology gap.

The tools are generally capable. The challenge is that most SMBs adopt them without mapping their existing sales workflow first. They plug in a lead scoring tool before their CRM data is clean. They set up automated sequences before they’ve agreed on what a qualified lead even looks like. The technology surfaces problems that were always there — it just makes them impossible to ignore.

A solid implementation follows a logical sequence:

• Audit first. Map the current sales process step by step. Where does time go? Where do leads stall? Where does information get lost between handoffs? This takes a few days but prevents weeks of wasted effort later.
• Clean the data. AI systems make decisions based on what they’re given. Incomplete CRM records, duplicate contacts, and inconsistent field entries all degrade output quality. Before deploying any AI layer, the underlying data needs to be in reasonable shape.
• Start narrow. Pick one use case — lead qualification, or pre-call research, or follow-up automation — and get that working well before expanding. Trying to automate everything at once is how implementations stall.
• Train the team on the why, not just the how. Salespeople adopt tools when they understand how those tools make their job better. Showing someone how to use a dashboard is less effective than showing them that it saved them ninety minutes last Tuesday.
• Measure against a baseline. Define what success looks like before you start. Conversion rate, time-to-first-response, number of qualified conversations per week — pick two or three metrics and track them from day one.

In our work helping founder-led professional services and recruitment firms build out AI-assisted sales workflows, the most common breakdown isn’t technical. It’s that the team didn’t agree on what a qualified lead looks like before they asked an AI to score them. That’s a conversation that needs to happen at the whiteboard, not in the software settings.

---

Key Concepts: A Working Vocabulary

If you’re evaluating AI sales tools for the first time, a few terms come up repeatedly. Here’s what they actually mean in practice.

Lead scoring — A system that assigns a numerical value to each prospect based on how likely they are to convert. AI-powered lead scoring uses behavioural and firmographic signals, rather than just demographic data, to rank prospects. Useful when you have enough historical data for the model to learn from.

Conversation intelligence — Technology that analyses recorded sales calls (with appropriate consent) to surface patterns: what objections come up most, where deals tend to stall, how top performers open calls differently. Typically used for coaching rather than automation.

Sales engagement platform — A tool that manages outreach sequences across email, phone, and sometimes LinkedIn. AI adds a layer of personalisation and timing optimisation on top of the basic sequence logic.

AI agent — An autonomous system that can take a defined set of actions on its own: sending a follow-up email, updating a CRM record, booking a meeting, routing a lead to the right salesperson. The agent operates within rules set by a human but doesn’t need to check in for each individual action.

Intent data — Signals that suggest a prospect is actively researching a solution in your category — for example, visiting competitor websites, reading relevant content, or engaging with specific topics on professional networks. Third-party intent data is available through several providers and can be used to time outreach more precisely.

---

Common Pitfalls to Avoid

Even well-resourced teams make predictable mistakes when deploying AI in sales. These are worth knowing before you start.

Over-automating early-stage outreach. Fully automated cold outreach at scale tends to erode deliverability and brand perception quickly. AI should assist personalisation, not replace the human judgment about whether to reach out at all.

Ignoring change management. A tool that the sales team doesn’t trust or doesn’t use delivers no return. Budget time for training, for answering questions honestly (including the ones about job security), and for gathering feedback from the people using the system daily.

Conflating activity with output. AI can dramatically increase the volume of outreach your team sends. More emails, more calls, more sequences. But if those activities aren’t generating qualified conversations, volume is just noise. Track outcomes, not just inputs.

Treating AI output as final. AI-generated prospect briefs, follow-up drafts, and lead scores are starting points. They require human review. A salesperson who sends an AI-generated message without reading it first is one hallucinated detail away from an embarrassing conversation.

Underestimating ongoing maintenance. AI systems need to be tuned over time. Lead scoring models drift as market conditions change. Sequence copy gets stale. The initial setup is not a one-time effort — budget for ongoing optimisation.

---

Realistic Expectations on ROI

It’s worth being direct here: the specific ROI figures that circulate in AI marketing content are almost always unreliable. Companies that report triple-digit ROI in their first year are typically measuring it in ways that don’t hold up to scrutiny, or they’re outliers presenting as norms.

What the research does support — from McKinsey’s annual State of AI reports and Forrester’s B2B sales surveys — is that sales teams using AI assistance consistently report meaningful productivity gains and improvements in pipeline quality over 12-month periods. The range varies significantly based on how well the implementation was done, how much the team actually adopted the tools, and how cleanly the underlying data was managed going in.

A realistic horizon for a founder-led SMB deploying AI in sales for the first time:

• 30 days: Initial time savings visible in research and follow-up tasks. Team still in adjustment period.
• 60 days: Lead quality data starts to reflect improved qualification. Sales cycle data begins to show movement.
• 90 days: Enough signal to evaluate whether the implementation is working and what needs adjustment.

Return on the investment in AI tools and implementation is real, but it compounds over months, not weeks. Teams that stick with the process and iterate consistently are the ones that report meaningful results.

---

Industries Where This Is Already Working

Certain SMB categories are seeing particularly strong results from AI-assisted sales, largely because their sales processes are well-defined and their deal sizes justify the investment.

Recruitment and staffing agencies deal with high volumes of both clients and candidates. AI helps prioritise outreach to hiring managers showing active signals, automate role-specific follow-up, and surface which open positions are most likely to close quickly.

Commercial real estate brokers benefit from AI-assisted prospect research — particularly around identifying tenants approaching lease renewals, buyers with relevant search activity, or investors in specific asset classes. The pre-call brief use case is especially strong here.

Legal and professional services firms often have long sales cycles with multiple stakeholders. AI helps track engagement across those stakeholders, flag when a deal has gone quiet, and draft follow-up communications that are specific to the stage of the relationship.

HVAC and trades businesses with a commercial sales function use AI primarily for follow-up automation and lead routing — making sure that inbound enquiries get a response fast, and that field technicians can flag upsell opportunities without creating manual CRM burden.

Accounting practices use AI to support business development by identifying clients approaching key financial events — year-end, funding rounds, M&A activity — that signal a need for additional services.

---

Where to Go From Here

AI won’t fix a broken sales process. But it will meaningfully improve a functional one — by reducing the friction that stops salespeople from spending time on what actually matters.

The technology is mature enough to deploy today for most SMB sales functions. The question is no longer whether to adopt it, but how to do it in a way that actually sticks.

If you’re a founder or sales leader trying to figure out where to start, the most useful first step is a clear-eyed assessment of where your current process loses time and loses deals. Everything else follows from that.

If you’d like to think through what that looks like for your business specifically, you can book a strategy conversation with the Basalt team here: https://cal.com/eliott-ardisson-kzq7zs/ai-strategy-call