The Analytics Stack Is About to Get a Lot More Like Your Content Stack
Every social media operator I know has hit the same wall. You’ve got the content calendar humming, the engagement metrics are climbing, and then your boss or your client asks for a “quick” look at how the last quarter’s TikTok strategy actually performed against the LinkedIn push. You open your analytics dashboard, and it’s a mess of disconnected charts, stale data, and a UI that was clearly designed by engineers who have never scheduled a post in their lives. You end up exporting CSVs, stitching together screenshots in Canva, and sending a PDF that’s outdated the moment you hit “send.”
This is the dirty secret of the creator economy: we are drowning in data but starving for insight. We have platform-native analytics that tell us something happened, but rarely why. We have third-party tools that aggregate metrics but often fail to give us the context we need to make a decision. And for anyone trying to build a sustainable business on the back of their content, this is a crisis. You can’t optimize what you can’t measure, and you can’t measure what you can’t see clearly.
That’s why the news that Evidence has open-sourced the core of its business intelligence platform—dubbed Evidence Core—isn’t just a story for data engineers. It’s a signal for anyone who runs social media like a serious business. The same principles that make a great content workflow—version control, modularity, and the ability to hand off tasks to an AI agent—are now being applied to the analytics layer. And that’s a shift worth paying attention to.
My take: the future of social media analytics isn’t a prettier dashboard; it’s a codebase. And Evidence Core just made that future a lot more accessible.
The Problem: Your Analytics Are a Black Box, Not a Workflow
Let’s be honest about the state of social media analytics. For most creators and small teams, the workflow looks something like this: you log into Instagram Insights, you screenshot the chart that shows a spike in profile visits, you paste it into a Notion doc, and you write a paragraph about how “the Reels strategy is working.” Then you do the same for YouTube Studio, then for X Analytics, then for LinkedIn. By the time you’ve compiled the report, you’ve spent an hour, and you have a document that’s already obsolete.
The tools that promise to solve this—the Buffers and Hootsuites of the world—are great for scheduling and basic reporting. But they’re often a black box. You get a graph of “engagement” but you can’t drill into the query that generated it. You can’t ask “what was the average watch time for videos posted between 6 PM and 9 PM on Thursdays?” without hoping the tool has a pre-built filter for it. And if you want to merge that data with your own sales numbers or your email list growth, you’re out of luck.
This is where the “business intelligence as code” philosophy hits home. The idea, as Evidence’s maker describes it, is to define your entire analytics platform—metrics, dashboards, reports—as code. For a social media operator, that means your analytics are no longer a separate, opaque tool. They become a part of your content pipeline, subject to the same version control, review, and automation as your video scripts and caption drafts.
I’ve been using a similar mental model for my own reporting for the last year. I don’t use a BI tool; I use a combination of Airtable and a Python script that pulls data from the APIs of Meta, TikTok, and YouTube. It’s clunky, but it gives me the one thing off-the-shelf tools can’t: the ability to ask a novel question and get an answer without waiting for a feature request. Evidence Core seems to formalize this approach, making it accessible to people who don’t want to write a custom scraper.
Why TikTok Creators Should Care More Than LinkedIn Ones
If you’re a LinkedIn influencer, your analytics are relatively simple: impressions, engagement rate, follower growth. The platform’s native dashboard is honestly fine for that. But if you’re a TikTok creator, the data is a firehose. You have video views, average watch time, completion rate, traffic source breakdowns, follower demographics, and a dozen other metrics that shift in importance depending on the algorithm’s mood. A static dashboard can’t keep up.
This is where the “code as analytics” model shines. You can write a query that says, “Show me the average watch time for videos longer than 60 seconds that were posted on a Friday, broken down by whether they used a trending sound or not.” That’s a question that could genuinely change your content strategy. But you can’t ask it in a standard dashboard. You need a tool that understands the structure of your data and lets you interrogate it dynamically. Evidence Core, with its SQL-and-markdown foundation, is built for exactly this kind of interrogation. It’s not about prettier charts; it’s about deeper questions.
How Evidence Core Differs from the Status Quo
To understand why this launch matters, you have to look at the incumbents. Looker and Tableau are powerful, but they’re enterprise-grade behemoths. They require dedicated data teams to maintain, and they’re priced for companies with six-figure software budgets. For a 10-person agency or a solo creator with a Patreon, they’re overkill and out of reach.
On the other end of the spectrum, you have tools like Metricool or Later, which are built for social media managers. They’re easy to use, but they’re also limited. They show you the metrics they want to show you, and they rarely let you connect your data to other sources. They’re like a content calendar that can only schedule to one platform—it’s better than nothing, but it’s not a system.
Evidence Core sits in a fascinating middle ground. It’s not a drag-and-drop dashboard tool. It’s a framework that assumes you have a data warehouse (or at least a database) and that you want to write SQL queries to pull the data you need. This is a big ask for a typical social media manager. But for an indie founder who’s comfortable with a little code, or for a growth marketer who’s tired of being limited by their tools, it’s a revelation.
The key differentiator, as Adam McAskill, co-founder of Evidence, notes in the launch comments, is the new CLI designed for coding agents. This is a massive deal. We’re seeing the rise of AI agents like Claude Code and GitHub Copilot that can write code, but they’re often useless without a clear framework to work within. Evidence Core gives these agents a structured environment where they can define metrics, build dashboards, and maintain reports. The maker even suggests that Claude Code can get you up and running quickly by simply prompting it to set up a new project.
This is where the “trustworthiness” flag goes up. The maker’s claim is promotional, but the underlying point is sound. In my experience, AI agents are only as good as the constraints you give them. A framework like this provides those constraints. It’s the difference between asking an intern to “make a report” and asking them to “fill out this template with the data from this spreadsheet.” The latter is more likely to produce something useful.
Where the Math Breaks
Let’s talk about the elephant in the room: the learning curve. Evidence Core is not for the faint of heart. It requires a working knowledge of SQL. You need to understand joins, aggregations, and window functions. If the last time you wrote a query was in a high school computer science class, this tool will not be your friend.
The source material is clear that the framework is for developers and data teams. The comments from early users are all from people who are building internal tools or data products. This is not a tool for a social media manager who just wants to see their weekly reach. It’s a tool for someone who is building the analytics infrastructure for that social media manager.
My take: if you’re a solo creator, this is probably overkill. But if you’re an agency that manages social for multiple clients, or a startup that wants to build a custom analytics dashboard for your audience, this is where the value lies. It’s the difference between buying a pre-built house and buying a set of blueprints. The blueprints are more work, but they let you build exactly what you need.
What Creators and Teams Can Borrow from This Philosophy
Even if you never touch Evidence Core, the philosophy behind it is worth stealing. Here’s what I’m taking away from this launch and applying to my own social media operations:
1. Version control for content strategy. The biggest problem with social media strategy is that it’s often a moving target. You change your posting schedule, you pivot your content pillars, and you forget why you made those changes. By treating your analytics as code, you get a version history. You can see that your engagement rate spiked in March, and you can look at the code to see exactly what metric definition you were using at the time. That’s powerful for retrospective analysis.
2. The agent is your junior analyst. I’ve been experimenting with using AI to draft my monthly performance reports. The results are mixed. Sometimes it’s brilliant; sometimes it’s a hallucination factory. The problem is that I ask it to “analyze my data” without giving it a structure. Evidence Core’s approach suggests a better way: give the agent a framework, a set of defined metrics, and a template for the report. Then let it do the grunt work. The maker’s comment about coding agents working productively with the tool is a lesson for all of us. The agent isn’t the strategist; it’s the analyst. You still have to define the questions.
3. Reporting as a product, not a chore. Most of us treat reporting as a necessary evil. We do it because clients demand it or because we need to justify our ad spend. But what if you treated your report as a product? What if you designed it to be beautiful, interactive, and useful? Evidence Core’s focus on dashboards and reports as code suggests a different mindset. The report isn’t a screenshot; it’s a living document that can be updated and re-run. For a social media team, this could mean a shift from monthly PDFs to a live dashboard that the whole team can check. That’s a game-changer for internal alignment.
The Repurposing Playbook, But for Data
We talk a lot about repurposing content—turning a YouTube video into a TikTok, a LinkedIn post into a thread. The same principle applies to analytics. You might have a core dashboard for your internal team, but you can also generate a client-facing version with different metrics. Or a public-facing version for your newsletter. With a code-based approach, you’re not rebuilding the report each time; you’re just changing the parameters. That’s efficiency that a drag-and-drop tool can’t match.
Where My Judgment Says It Falls Short
I’m excited about this launch, but I’m also a realist. Here’s where I think the hype needs to be tempered.
The “open-source” caveat. The source material says they are open-sourcing the framework that powers their commercial product, Evidence Studio. This is a classic open-core model. It’s great for adoption, but it also means the best features are likely to be in the paid version. The open-source version will be a solid foundation, but you’ll probably need to pay for the more advanced features like managed hosting or premium support. That’s not a criticism; it’s just a business model. Just be aware that “free” here means “free to get started,” not “free forever.”
The data warehouse prerequisite. To use a tool like this effectively, you need your data in a centralized location. That means setting up a data pipeline from your social platforms, which is a non-trivial task. The APIs for Instagram, TikTok, and LinkedIn are not always the most straightforward to work with. You’ll likely need a tool like Fivetran or Airbyte to get the data into a warehouse like Snowflake or BigQuery. This is a significant investment in time and money. If you’re not prepared to make that investment, this tool isn’t for you.
The UI is still an afterthought. The product is built for developers, and it shows. The focus is on the code, not on the visual design. For a social media manager who is used to the slick interfaces of Canva and CapCut, the output might feel a bit barebones. You can customize it, but it requires effort. If you’re looking for a beautiful out-of-the-box dashboard, this isn’t it.
The agent is only as good as the data. The maker’s comment about Claude Code getting you up and running is optimistic. In my experience, AI agents are great at writing code, but they’re terrible at understanding business context. They can generate a query that calculates the correct metric, but they can’t tell you why that metric matters. You still need to be the strategist. The agent is a tool, not a replacement for your judgment.
What I’d Watch / Test Next
This week, I’m not going to rewrite my entire analytics stack. But I am going to take a few concrete steps based on this launch.
First, I’m going to read the Evidence docs. Even if I don’t use the tool, the documentation will likely contain useful patterns for how to structure metrics and reports. It’s like reading a cookbook—you don’t have to cook every recipe, but you’ll learn techniques that improve your own cooking.
Second, I’m going to prototype a single report. I’m going to pick one platform (probably YouTube) and one question (“What is the average view duration for my videos, broken down by topic?”) and see if I can build a simple report using Evidence Core. The goal isn’t to replace my current workflow; it’s to understand the ergonomics of the tool. Is it intuitive? Does the CLI make sense? Can I get an AI agent to help me with the boilerplate? This is a low-risk experiment that will tell me a lot about whether this approach is viable for my workflow.
Third, I’m going to talk to my data-savvy friends. I know a few engineers who are into analytics. I’m going to share the launch link and ask them what they think. The social media world is often siloed from the engineering world, but this launch is a bridge. I want to hear from people who live in SQL every day. Their perspective will be more valuable than any influencer’s hot take.
The bottom line is this: the tools we use to measure our impact are becoming as important as the tools we use to create our content. For too long, analytics has been an afterthought. Evidence Core is a reminder that it should be a core part of the workflow. It’s not a tool for everyone, but it’s a tool that signals a shift. The future of social media analytics is not a prettier dashboard; it’s a more thoughtful, more integrated, and more code-driven approach to understanding what works. And that’s a future I’m ready to explore.
