Start with the friction you already know
Turn7's client portal started with a common agency problem: getting a proposal into a prospective client's hands quickly.
The team had tried existing proposal software, but moving project details into those tools took time. Once Turn7 had worked through scope, pricing, timelines, and features, Erick wanted a faster way to turn that information into something a client could sign.
“When we get a lead, we want to move quick on it,” Erick explained. “It's about the response time.”
So Turn7 built its own portal on WordPress.
The portal can turn the team's project details into a proposal within minutes, complete with costs, timelines, a custom cover image, and signing functionality. Once a client signs, it kicks off other parts of Turn7's workflow, including creating project management tasks in Asana, generating invoices, and sending confirmation emails.
WordPress gave the team much of the infrastructure it needed, including user logins and ways to keep account information private. That allowed Turn7 to focus on connecting its platforms and creating a front-end experience that was easy for clients to use.
That client experience has become an important part of the portal, too. Each proposal is branded and personalized, helping Turn7 establish confidence before the project begins. Erick told Cory that the agency wants to “start off with high trust from the get-go.”
The portal has continued to evolve since its first version in 2018, with client feedback helping guide what comes next. Some clients have even returned to the portal expecting to find ways to submit maintenance requests or manage their websites.
Those features don't exist yet, but the behavior has given Turn7 another signal about where the portal could go. As Erick explained, feedback helps the team make sure they're “building the right features and not spending time on things that we might not necessarily need.”
Find the places AI can make the work better
The portal is only one part of how Turn7 is using AI across the agency.
During technical planning, the team uses tools including GitHub Copilot to explore different approaches to a project and consider questions around architecture, hosting, scalability, and long-term maintenance.
Erick also asks AI to challenge the team's assumptions. One of his questions is simply: What are the cons of going with this?
That can expose tradeoffs before Turn7 makes a recommendation. A less expensive approach might have limitations around scalability, for example. If the client doesn't need that scale, it could still be the right choice. The team has more information to make that decision and explain it to the client.
Turn7 is taking a different approach with work that's more predictable.
For one restaurant client, the agency has an AI agent that creates and deploys employee email signatures. The agent already has instructions for how to build them, so it only needs information such as the person's name, role, and email address.
That means a project manager can request a new signature without handing the task to a developer unless something goes wrong.
Turn7 is applying the same thinking to invoicing and other internal processes. The goal isn't to automate everything the team does. It's to identify work with a clear process and decide where AI can make that process faster or easier.
Follow where client needs are changing
Turn7's experiments with AI are also opening the door to different kinds of client work.
One recent project had little to do with building a website. A client had accumulated roughly 40,000 emails in an inbox, including newsletter replies, subscriptions, and messages requiring a response.
Turn7 is using AI to sort and label those messages, identify the ones that need attention, and archive or delete others based on their category.
Clients using AI on their own doesn't necessarily mean losing their business, either. Erick has already seen how AI can change the role an agency plays in a website project, while still leaving plenty of room for its expertise.
A longtime client asked Turn7 for a proposal for a new WordPress website. Instead of moving forward with the build, the client took the project specifications and used AI tools to create a version of the site over a weekend.
Then the client came back to Turn7 for help deploying it and working on SEO.
For Erick, experiences like this are changing where the agency's expertise can enter the process. Clients may be able to build more on their own, but questions around security, scalability, integrations, marketing and long-term maintenance remain.
Erick describes the opportunity as helping clients move from "40 to 80." A client might use AI to get a project functional, but unpolished, roughly 40 percent done. Turn7's job is closing the gap to the other 80 percent: the work that makes a site secure, findable, and built to last. AI may help a client get a site built and running on their own, but getting it further means addressing areas like SEO, security, scalability, integrations, and marketing. That's where Turn7 can bring its experience back into the project.
Rather than seeing accessible AI tools as something his team should resist, Erick wants Turn7 to figure out where its expertise becomes most useful. A client might arrive with something they've already built. Another might bring a business problem, like 40,000 unsorted emails, that doesn't fit into the agency's traditional services.
Both create opportunities to start with the client's problem and determine where the agency can help next.
Build what earns its place
Turn7 has plenty of ideas for where AI could fit next. The harder part is deciding which ones deserve the team's attention.
Erick's filter is practical: “What's the next one that's going to give us the best ROI or a better client experience?”
It's also a useful starting point for other agencies. Before adding another AI tool or building another automation, look at the work your team already does. Where are you losing time? What does your team repeat from project to project? Where are clients asking for something you could deliver more efficiently?
Pick one of those problems, define what a better outcome would look like, and then see where AI can help.
That's the pattern behind much of what Turn7 has built. The portal started with a slow proposal process. Other automations grew from repetitive work. New services are emerging from problems clients are already bringing to the agency.
You don't need to find every place AI could fit into your agency. Find one place where it can make the work better, prove its value, and build from there.