GKV's two-person analytics team uses IQ Agents to streamline reporting, automate recurring tasks and close the gap between data and strategic recommendations.
The case
Anna Alexiadi Smyrniotaki runs a two-person analytics team at GKV covering around 12 clients. Their mission is to give client teams the data-driven insights and recommendations they need to make the best possible decisions for clients. But reporting took up most of their time.
The team produces monthly insights reports for 12 of their clients. Each report took 10 to 15 hours, which adds up to 120 to 180 hours reporting time every month.
Both analysts spent three out of every four weeks on monthly reporting, and some clients didn't get last month's report until the final week of the month.
That leaves only one week in a month for providing data-driven insights and recommendations to client teams.
Whatagraph IQ chat and summary writing had already cut reporting time by 40%. But the recurring work still landed on the team every month.
So when IQ Agents opened for early access, Anna had a plan: hand the recurring groundwork to agents, and use the time to focus on data analysis and strategic work for client teams.
“If the agents can complete monthly reporting in the first one or two weeks of the month, that could exponentially save us time. Then we have the rest of the month left to connect with stakeholders, answer ad-hoc questions, and do the more human and strategic parts of the job.”
The solution
Co-creating reports with client specialist agents
Anna decided to build a specialist agent for each client that drafts monthly reports for each client on a schedule and analyzes key insights, with her in the loop at every stage.
To start off, she described what she wanted to build in Whatagraph IQ Agents:
The agent would pull the data from the governed data foundation the GKV team had built in Whatagraph:
- The client's folder containing all the previous reports
- Every connected data source
- The client's source group (that combines data from multiple sources into one group) and campaign dimensions, so cross-channel numbers roll up the way GKV defines them
- The client's theme and color palette, applied to every report, tab and widget the agent creates or edits
Anna designed each agent as a co-worker for her team that drafts:
- Monthly reports. The agent drafts the client's performance report on a set schedule, pulling the month's data from every connected source. Anna's team reviews it, pushes it deeper, and signs it off before it reaches the client.
- Narratives. The agent drafts the wins, watchouts and recommendations into the report widgets. Anna's team reviews everything and adds the context only they have about each client.
- Ad hoc answers. The agent runs the first pass when a client or account lead asks something. Anna checks the answer, goes back and forth, and sends a more in-depth analysis to the client.
Anna built the first agent on August 18th and scheduled it to draft a monthly performance report on the 2nd of every month.
On September 2nd, the draft was ready, commentary and insights included. The agent took about seven minutes to write it because of the data foundation and previous reports Anna had built on Whatagraph.
It also read the previous reports in the client's folder and wrote to the same shape, which Anna hadn't asked it to do.
“I have a wealth of prior reporting available in Whatagraph, and so the agent was able to, without even me telling it to, match the output to what the previous reports looked like. That was a surprise."
Then the co-creation started. Anna read the report and the narratives, then sent follow-ups to push the agent deeper and recheck things.
“I went back and forth with the agent, hey, dive deeper into this, check your numbers here, be better here. With that co-creation, the output is better."
Creating a client-wide digest agent
Specialist agents work for each brand, but Anna also wanted an overview of all her clients without having to open 15 different reports.
So, she built a digest agent that would audit creatives and media across all of GKV's clients every day. This means Anna finds out what moved across all the clients before they ask and can take proactive action.
“I created an alert system, a little bit like a digest system that tells me, hey, here are the big movers and shakers for all accounts, and that was very useful."
The agent runs every day at 09:00 without Anna having to open the laptop or log into Whatagraph.
Once she logs into work, the audit report is already there for her. From there, Anna makes the judgment calls: which movements matter, what needs a closer look, and which campaigns to flag to the media team for optimization.
The results
Streamlined reporting: 70+ hours back every month
A regular monthly report used to take 10 to 15 hours. With an agent drafting it, the same report can go out in one to two hours when there's no extra exploration or manual data work involved.
On a typical report, Anna estimates it takes around four hours with agents, her review included.
At that rate, the 12 regular reports take about 48 hours a month, down from 120 to 180. That's roughly 70 to 130 hours back every month.
“IQ agents enable us to get started on monthly reporting from a rough draft of performance takeaways, so we can focus on curating deeper insights and more actionable recommendations."
Anna credits two things the team had already set up in Whatagraph:
- The data foundation. Every source for every client is connected in one place and kept up to date.
- Their existing reporting. The agents build on the dashboards and reports the team had already created.
“The agents are both extending and leveraging the depth of the Whatagraph platform and all of the work that went into it.”
Delivering proactive, data-driven insights for client teams
With the time saved on reporting, Anna and the team now have the space for deeper analysis and proactive monitoring for client teams.
For instance, when Google Ads changed how it optimizes target CPA campaigns, Anna wanted to know which of their clients were affected.
Before agents, that meant going into Google Ads or building a widget in Whatagraph that listed every campaign by objective, then working out the scope from there. This would take a couple hours.
Instead, she asked the client specialist agent. A few minutes later she had the answer and sent it to the client team for optimization.
This wasn't about time saved. This was net-new work agents freed up time for.
“It's one of those tasks that would have taken me a couple of hours, but because I have the agents, I can just ask the question and see what comes up."
Answers to client questions now go deeper too. When a client asked about their TikTok and Google Analytics performance, Anna went several layers deeper than she'd usually have time for and delivered deeper analysis to the client.
“I'm able to get back to the client not only faster, but with more informed feedback."
Follow-up questions and new ad hoc analyses now start in about 10 minutes. Analyses that used to need at least a two-day turnaround go out in one to two hours, up to 92% faster.
The analysis, the judgment and every recommendation that reaches a client still sit with Anna and her teammate. The agents handle the groundwork, so the team can spend its time on the thinking.
“We believe the IQ Agents advance our team's mission to empower the client teams with data-driven insights and recommendations that ensure the best possible decisions and outcomes for our clients."