Claude Sonnet 5 Just Launched. Here’s How a Revenue Leader Would Actually Use It.

Anthropic released Claude Sonnet 5 on June 30, and it’s already the default model across Claude’s free and paid plans. The specs are worth knowing: a one-million-token context window, noticeably stronger performance on multi-step work, and pricing well under Anthropic’s flagship Opus model. But the spec sheet isn’t really the story. The story is that early testers keep saying the same thing: it finishes the task instead of stalling halfway through. That’s the piece that’s kept most agentic AI stuck in pilot mode for the last couple years.

I’ve spent 20+ years in marketing and revenue leadership across healthcare and SaaS, and I’ve watched a lot of “game-changing” tools come and go. Most of them just help you do a task faster. Very few actually change how you’d approach the work itself. I think this one might be the second kind, so I want to walk through how I’d actually use it, function by function, the way I think about the business.

Where this fits, role by role

If you’ve run a marketing org at a mid-size SaaS or healthcare company, the shape is pretty familiar no matter the industry. Someone owns social, events, and trade shows. Someone owns content and campaigns. Someone lives in the field, pulling signal from sales calls and customers, often through a tool like Crayon. Someone owns intent and pipeline data, usually out of HubSpot, with Clay layered in for enrichment. And distribution runs across Instagram and LinkedIn, with TikTok now showing up in enterprise strategy too, not just consumer brands.

Here’s where I think Sonnet 5 actually earns a seat on that team, function by function.

Social, events, and trade shows live and die on logistics and repetition, booth messaging, follow-up sequences, sponsor recaps, and one session turned into five different post formats. A longer context window means you can hand it an entire event’s worth of materials at once and get something that actually sounds consistent, instead of re-explaining your voice every single time.

Content and campaign work is where the “finishes the task” part matters most. It’s not about writing faster. It’s drafting the brief, building out the variations, mapping them to the right funnel stage, and checking its own work before it lands on your desk.

Voice of the field is honestly the one I’d be most excited about if I were building this function today. Feed it call transcripts, win-loss notes, and Crayon’s competitive intel, and ask it to identify themes rather than just summarize. Being able to hold that much at once is what lets you catch patterns across a full quarter of sales conversations, rather than the handful you happen to remember.

Intent and pipeline work is where HubSpot and Clay already give you the raw data. What changes is going from a list of high-intent accounts to an actual drafted, sequenced outreach plan built off what you know about each one, with a person still deciding what goes out the door.

And distribution across Instagram, LinkedIn, and TikTok gets easier when the model can keep your brand guidelines, past performance, and each platform’s norms in mind at once, rather than treating every post as a blank page. More enterprise brands are testing TikTok now, and not because it’s trendy, it’s because the buyers have aged into it.

None of this is a SaaS-only story either. A services firm has the same event and content rhythm, just with longer sales cycles and more relationship-driven field intel. CPG brands live and die on social and retail media execution at volume, which is exactly the kind of repeatable, multi-step work this model was built for.

Where I’d actually put the budget: RevOps and sales enablement

This is the part most people skip when they talk about AI and marketing, and it’s the part I think matters most. Content production isn’t where the leverage is. RevOps and enablement is, the work that decides whether any of that content actually turns into pipeline.

Think about what those teams spend their time on. Keeping playbooks current. Making sure reps have the right competitive positioning in the moment they need it. Tracking which content is actually influencing closed revenue. Turning field feedback into real messaging changes. That’s a lot of pulling together scattered information, and it’s exactly where a bigger context window and better follow-through help most. When a model can hold your CRM data, your win-loss history, and your current messaging at the same time, and actually make the updates instead of just describing them, it starts pulling real weight on the ops side, not just the writing side.

The question every CMO is actually getting asked

Underneath all of this is CAC and CPL. That’s the real conversation happening in boardrooms right now: how does AI actually bring these numbers down, not just make the team feel busier?

The honest answer isn’t content volume. Publishing more, faster, doesn’t lower CAC if none of it is findable or differentiated. What moves the number is waste going down and conversion going up on the spend you’re already making. Sharper intent data means fewer dollars chasing the wrong accounts. Faster follow-up on high-intent signals means fewer leads going cold before sales ever gets to them. A RevOps function that keeps messaging current means fewer campaigns built on stale positioning burning budget.

There’s a bigger shift sitting behind all of that, too. Buyers are starting to get recommendations from AI assistants before they ever open a search engine, and that’s rewriting what “getting found” even means, on top of everything Google’s own algorithm changes have already done to organic. That’s a big enough topic to deserve its own post. If visibility is changing this much, measurability has to change with it. More on that soon.

And for the SDR team specifically

SDRs are where all of this either becomes pipeline or doesn’t, so they deserve their own callout. The daily grind for most SDR teams is research, personalization, and volume, enough accounts prospected, enough relevant outreach written, fast enough follow-up that a warm signal doesn’t go cold before anyone responds.

That’s exactly the kind of workload that benefits from a model that can hold real context and actually finish a task rather than draft one message and stop. Feed it the intent signals coming out of HubSpot and Clay, the account history, and your current messaging, and it can build an outreach sequence that’s genuinely personalized instead of a template with a first name swapped in. It can also handle the unglamorous stuff that gets skipped under quota pressure, pulling recent news or trigger events on an account before a rep ever picks up the phone.

None of that replaces a rep’s judgment on tone, timing, or reading how a prospect responds. What it does is give reps back more of their day for the actual selling conversation, rather than the research that has to happen first. For a manager, that changes coaching too. Less time checking whether the research got done, more time coaching the conversation itself.

What I’d actually do with it

Don’t hand this to one person and call it “the AI thing.” Every function touches different data and needs different output, so the value shows up when you use it across the business, not just in one corner of it. The leaders getting real value here aren’t the ones using it to write faster. They’re the ones who’ve figured out how to give it enough context, and enough judgment, to actually finish the work.

The model doing more of the task doesn’t mean you need less strategy. It means your strategy is finally worth more because there’s now a real way to execute it at the pace you always wanted.

If you’re thinking through how to put frameworks like this to work inside your organization, I’d love to talk. Reach out at robo@robinbowling.com or visit www.robinbowling.com.