My past challenges in data retrieval

While working on Ads Planning tools at Google, I frequently encountered requests to extract data showcasing the consistency and accuracy of our forecasting tools. This process drove me to my wit's end in two distinct phases.

Phase 1: I either had to craft new Dremel queries from the ground up or modify existing ones. This often meant context-switching and revisiting past work to understand it anew.

Phase 2: As the volume of these data requests surged, I created a suite of visualization dashboards. These were designed to assist engineers and PMs in monitoring the performance of the forecasting tools. However, new challenges emerged:

• Often, those who requested the data didn’t even access these dashboards.
• When they did, a third of the time, the data pipeline supporting the dashboards malfunctioned, hindering data visualization. Surprisingly, this is a common occurrence at Google.
• If the dashboards and data pipelines functioned correctly, I'd frequently receive follow-up questions. These would lead me to another unique data extraction task, potentially resulting in yet another dashboard addition. Over time, the sheer volume of dashboards became overwhelming, even for me.

From these experiences, I distilled several guiding principles for my next venture:

• Simplicity: Planners should access the required information with minimal clicks, scrolls, and searches.
• Low latency: Planners don’t want prolonged waits; they desire immediate data, especially during brainstorming sessions.
• Accuracy: Only precise data can guide optimal decisions.

Market solutions fall short of expectations

Solutions already existed for marketers, with Supermetrics being a notable example. While it's a commendable extension for Google Sheets, its interface can be cumbersome, especially when managing multiple accounts. Additionally, channeling marketing data through Supermetrics to Big Query and Looker Studio proved costly and reintroduced previous challenges.

I want to build a better solution

Consequently, I embarked on a mission to craft a Google Search-esque experience for marketers. My goal is to enable users to achieve "understanding without engineering." Just as Google Search eliminates library visits, my tool aims to reduce the need to juggle between platforms like Google Ads, Sheets, and Looker Studio.

Building on the aforementioned principles, I also incorporated the beloved Costco shopping experience:

• Affordable pricing: I aim to achieve this by indie hacking, steering clear of VC or PE investments.
• Stellar customer service: No diplomacy. No upselling. It makes me happy knowing that I deliver real value to customers through my software.

I’ve been building Adzviser, a ChatGPT plugin that allows you to connect to multiple data sources such as Google Ads, Meta Ads and Instagram, etc. With a prompt such as “How much did I spend on Google Ads last month and this month”, ChatGPT returns you the data. I proudly launched the initial version in late August, 2023.

If my journey resonates with you or you're curious about Adzviser, feel free to reach out at zeyuan.gu@adzviser.com or connect with me on Twitter.