No items found.
Right side up logoRight side up logo

RSU

Strategy

RSU

Strategy

Jump to section →

On this page

Inside an AI-Forward Marketing Company: How We’re Putting AI to Work

Published

November 17, 2026

Updated

[tldr]

TL;DR: How we're putting AI to work at Right Side Up

  • AI adoption takes more than access to tools. At Right Side Up, we focus on training, incentivizing experimentation, and holding teams accountable for meaningful outcomes.
  • Sharing is how individual wins become company-wide improvements. Internal show-and-tells and shared resources help our teams learn from each other and scale successful workflows.
  • Our workflows come in all shapes and sizes. From client reporting that takes minutes instead of hours to automated product release communications, our internal teams have built all sorts of high-impact workflows.
  • Human expertise remains essential. AI handles repetitive work and information gathering, while marketers provide context and judgment.
  • Start with problems worth solving. Look for frequent, fragmented, or judgment-heavy tasks, then build the workflows that make the work better.

[/tldr]

Almost every company claims to be AI-forward, but how many can actually back it up? 

We’d wager that number to be a lot lower. 

At Right Side Up, we seek that level of fluency from every marketer we bring into our talent collective, which is why we developed a thorough AI vetting process. We then deploy those AI-fluent marketers through our AI Services to help clients identify, build, and adopt better ways of working with AI.

But it’s not just about our talent pool. We also hold our internal teams to the same standard, and understand it requires much more than just giving people access to ChatGPT or Claude. 

Instead, we focus on deeply and systematically incorporating AI into our work. It requires a much more intentional approach but also significantly yields better outcomes—both for individuals and the team as a whole.

We’ve built an environment where our teams are encouraged to experiment, share what works, build repeatable systems, and prove that AI is actually making the work better.

Here’s a little more about what we’ve implemented—and some of our top learnings that you can apply to make your team (or even just yourself) more AI-fluent.

Building an AI-forward company starts with the environment

AI adoption can only happen when people have the skills, incentives, and space to experiment—and when there’s a system for spreading the best ideas beyond whoever discovered them first.

At Right Side Up, we’ve focused on three pieces: making sure people are trained, encouraged, and accountable.

We train people to get better at AI

AI education is an ongoing focus at Right Side Up, led by a cross-functional taskforce made up of our smartest AI minds.

First, all employees are required to participate in what we call “Foundations” classes. This includes getting familiar with the basics of how we approach AI as an organization as well as tooling, but also a deeper dive into the core platforms we use (including ChatGPT and Claude).

We also have weekly office hours where anyone can brainstorm, discuss, or troubleshoot AI ideas. These are open to all fluency levels.

On the async front, our “show and tell” Slack channel, which we’ll discuss more in the next section, allows teams to get inspiration and education on what others are doing with AI internally. We 

And last but not least, everything is neatly documented in an ever-growing wiki, where anyone can learn about our tools and processes—plus get recordings of any AI training session or office hours we’ve ever hosted.

We make experimentation visible—and worth sharing

A lot of useful AI experimentation starts informally, like someone finding a more organized way to prep for client calls or another teammate realizing that a repetitive analysis could be turned into a reusable workflow.

The problem is that, often, these discoveries stay trapped inside one person’s chat history.

That’s why it’s important to encourage and incentivize sharing wins, and why we created an internal AI “show and tell” Slack channel. In that space, employees share workflows, tools, prototypes, small wins, and lessons from things they’ve built. 

Our "AI Show and Tell" channel, where employees share AI processes, wins, and learnings.

Some are sophisticated, others delightfully simple—the whole point is about continuously learning and improving.

A workflow that saves one person 15 minutes a day is useful. But when that person shares it with the company, someone on another team may realize the same idea can solve a much larger problem.

We also actively reward experimentation, with quarterly prizes for the best projects submitted to the Slack channel. 

We measure whether AI is actually improving the work

Unmeasured AI experimentation can quickly become playing with shiny objects. Which, to be fair, is very fun, but not necessarily productive.

So AI adoption also shows up in how we set goals and evaluate progress, with each team and contributor having tailored objectives.

We don’t focus so much on superficial metrics like how many prompts someone sends or how many AI tools they’ve tried. We care much more about outcomes like:

  • Did we remove meaningful manual work?
  • Did we improve the quality or consistency of an output?
  • Did we make an important decision faster?
  • Did we create a reusable capability?
  • Did other people adopt the workflow?
  • Did the process ultimately perform better than what came before it?

Put simply, the measurement mindset shift that should be applied by every org is to go from “Are we using AI?” to “Is AI making us better?”

What our teams are actually building with AI

Culture, training, and goals help set the tone for AI enablement at Right Side Up. 

But now let’s get to the most exciting part of the whole deal: the projects themselves (and the results they yield).

Across Right Side Up, teams are using AI in very different ways, from lightweight automations that eliminate repetitive tasks to more complex systems that synthesize data, apply expert judgment, and support client-facing decisions.

The examples below, which come from various internal teams, illustrate that range. And while the specifics vary, the pattern behind strong AI use cases is consistent: 

Start with a real problem, build around the way the work actually happens, keep expert judgment where it matters, and measure whether the new approach is better.

Use case #1: Right Percent’s reporting bot

The problem:

Right Side Up’s B2B advertising arm, Right Percent, was struggling with data—and answers—living in too many places.

To answer a simple client question, someone had to log into several platforms, pull reports, dig through call notes, and stitch it all together by hand. That was slow, the results varied depending on who did the digging, and it didn’t scale as clients were added.

What they built:

The RP Reporting Bot pulls ad platform data, Fathom meeting notes, and client CRM or reporting data into one place. Then anyone on the team can ask it questions like “How did Meta perform last month compared to what we discussed on the last client call?”, and get an answer or a report back in seconds.

Where humans come in:

The bot handles the gathering, but people handle the thinking. The team still asks the right questions, checks the outputs, interprets what the numbers mean for each client—before deciding what to do next. The bot just takes the tedious data wrangling off their plate.

The result:

Pulling together a full picture of a client used to take hours across multiple platforms. Now it takes minutes, freeing up Right Percent’s client services team to spend more time on strategy and insight.

[tldr]

Words from the builder

“Writing clear prompts, giving the AI the right context, and connecting the right data sources made the difference between a novelty and a tool the team actually uses.”

[/tldr]

Use case #2: A revamped flow for product release updates

The problem:

Our lean product team ships changes every week, but the rest of the company didn’t always know what changed or how to use it. Writing release notes meant combing through Linear by hand, deciding what mattered to non-technical teams, writing how-to docs, and then reformatting the same update for Slack, email, and Slite. It was the kind of work that slipped when the team got busy, which is exactly when the most was shipping.

What they built:

A Claude workflow that runs as a scheduled task every Friday morning. 

What goes in: everything the team moved to a shipped state in Linear since the last digest, plus the issues we’ve tagged as worth announcing. 

What AI does: it audits the shipped work (including catching tickets that were closed but never actually built, and shipped features nobody tagged), checks whether each feature has a Knowledge Portal article in Slite and drafts one if it doesn’t, then writes the update once and adapts it for three channels. 

What comes out: a release notes page in Slite, an on-brand email draft in Gmail, and a Slack message draft, each linking to the how-to guides. Main tools: Claude (scheduled tasks plus custom skills for our brand, voice, and Knowledge Portal format), Linear, Slite, Gmail, and Slack.

Where humans come in:

Deciding what makes the cut. Not every shipped ticket is worth announcing, and sometimes the digest is narrowed to one or two headline features. Every draft is human-reviewed, and nothing goes out without final approval.

The result:

Release communication became a repeatable process instead of something that happened when someone had time. Humans now own editing instead of writing: reviewing the shortlist, tweaking the drafts, hitting send. And every announced feature now comes with documentation, so teams can actually use what we ship without having to ask how it works.

[tldr]

Words from the builder

"Write down every rule you’d explain to a new teammate (what counts as shipped, who the audience is, what the tone should be), because that playbook is what makes the output trustworthy week after week. And keep a human approval step for anything going to the whole company."

[/tldr]

Our talent is AI-vetted too 

We also prioritize AI fluency when it comes to the fractional talent we deploy.

Our fractional marketers have leveraged AI to build custom media mix models, reporting systems, creative-analysis workflows, research agents, ecommerce tools, campaign-audit systems, and more.

As we explain in our guide to vetting marketers for true AI fluency, knowing a few AI tools isn’t enough. We look for senior marketers who can pair deep domain expertise with systems thinking, technical fluency, and the judgment to know when AI should (or shouldn’t) be involved.

Because AI skill is impossible to measure using a single technique or metric, we combine several sources of evidence, including:

  • Onboarding data about how marketers use AI and the systems they’ve built
  • Real-world experience showing where AI has contributed to actual marketing outcomes
  • A live technical screen that evaluates tool depth, workflow design, AI limitations, human oversight, and marketing judgment

Want help identifying the right AI opportunities—or access to marketers who can build them? Talk to Right Side Up.

How marketing teams can become AI-operational

Most—if not all—of the experiments above didn’t come from brainstorming every possible AI opportunity, but from work that was already happening.

Our Director of Agentic Commerce, Jack Atlasov, recommends looking for tasks that are:

  • Repeated frequently: Your team keeps solving essentially the same problem.
  • Judgment-heavy: Doing the work well requires expertise or context that lives inside people’s heads.
  • Fragmented: Context has to survive across people, tools, or sessions.
  • Measurable: You can define what a better version of the process would look like.

[tldr]

AI resources for your marketing team

For a deeper dive, read our guide to finding the highest-value AI opportunities on your marketing team and watch our webinars on building reusable AI systems and becoming an AI-operational marketing organization.

[/tldr]

From there, the path to becoming more AI-operational is relatively straightforward:

1. Give the team direction

Leadership needs to make AI improvement an actual priority, not a side project people pursue when they have extra time. (Who has extra time these days anyways?) Right Side Up’s three levers—training, incentivization, and accountability—come in handy here.

2. Make sure the right context and data are available

AI is only as useful as the information it can work with. Teams need to think about where critical data, instructions, examples, and institutional knowledge live—and whether AI can access them effectively.

3. Turn the strongest opportunities into workflows

Start small. Prototype. Test. Define what “better” means before over-engineering the solution, and always keep humans involved where judgment is important.

4. Enable people to use what you build

Arguably the most important part of a workflow is whether or not people actually use it. Enablement comes in the form of training, documentation, ownership, and feedback.

5. Measure, improve, and share

Compare the new process with its predecessor. Keep what works, fix what doesn’t, and spread useful approaches across the team.

[tldr]

Building an AI-forward marketing team takes more than good intentions

The process of becoming more AI-forward is simpler in theory than execution. In real life, it requires a combination of AI fluency, deep marketing expertise, implementation bandwidth, and change management.

That’s what our AI Services are designed to provide.

Depending on where your team is today, Right Side Up can help:

  • Evaluate: Assess how your team currently works, identify high-impact AI opportunities, and prioritize what’s worth pursuing.
  • Build: Adapt an existing workflow when there’s a strong fit or build something completely new around your needs.
  • Train and scale: Help your team adopt the workflow, document it, and keep improving it.

And importantly, that work is led by senior marketers who understand both the marketing problem and the AI systems that can improve it.

See how Right Side Up can help your marketing team put AI to work →

[/tldr]

Vincent is a senior content marketer who finds equal joy in crafting in-depth guides and penning punchy subject lines. Before joining Right Side Up, he honed his skills in the fintech, insurance, and travel worlds—both agency-side and in-house. In his spare time, you can find him riding his bike or petting his cats.

FAQ

No items found.

Let's talk growth

Get in touch

Let's talk growth

Get in touch

Let's talk growth

Get in touch

Hungry for growth?

Sign up for growth marketing insights delivered straight to your inbox