Exactly. That "three PDFs stuck together" analogy is painfully accurate. I've seen teams waste more time reconciling static reports than they'd spend just querying the database directly.
Your click-filtering test is the right litmus test. But I'd push it further: even if Gauge supports basic dashboard-wide filters, you need to verify drill-down depth. Can you click a quarterly spike, then drill to the specific campaign, then to the individual asset? Or does it just filter to "Q1 content" and stop?
Many tools that market interactivity actually create new, isolated queries on drill-down, breaking the weighted average calculations mentioned earlier. So you get filtered but statistically meaningless data.
The detailed notes on Profound's three-panel structure are a useful start for the comparison, but you've only described its static output. For a real side-by-side test, you need to perform the exact interactive actions that the team would use daily.
Specifically, after seeing the "Top Content Q3" table, did you try clicking on a high-performing blog title to see if the 12-month trend line filtered to reflect just that piece's history? If the panels are truly separate, as others have noted, that click won't do anything, and you're just looking at a report. The value of a dashboard is in that connective tissue. Without demonstrating whether those panels communicate, the test is incomplete.
Let's keep it constructive
Yeah, that missing A/B test section is a huge oversight. If they can't even include a requested module, it makes you wonder what else they'd just decide to drop post-sale.
On the table point, you're spot on. A single column forces manual aggregation, which defeats the purpose. A real dashboard for content types needs at least a simple bar chart comparing average performance, with the ability to click and see the underlying pieces. Anything less is just a data dump.
Happy testing!
Yeah, that's exactly what I was wondering too. Having a column for content type in the table doesn't really show which one is winning overall. You'd have to manually sum or average that column yourself, which defeats the purpose.
It made me think about how you'd even build that aggregate comparison. Would you need a separate data pipeline to pre-calculate the average performance for each content type, or can these tools actually do that grouping on the fly from the raw data? That seems like a basic requirement.
rookie
So Profound only gave you the three panels you listed in your notes? That seems thin for a brief that asked for four things, especially the missing A/B test section. Did they say why it was left out, or was it just not part of their template?
The biggest red flag for me is that your brief had four clear requests, and Profound delivered three static panels. It's not just about the missing A/B test section, it's about the pattern. It means the tool is template-driven, not logic-driven. They took your prompt and shoehorned it into their pre-built "content snapshot" module.
You asked for a "clear indicator of which content type is performing best." A table with a "Content Type" column is a data list, not an indicator. An indicator implies a calculated aggregate, like a bar chart comparing the average conversion rate per type. The fact they didn't generate that tells you their calculation engine is likely very weak. You'll be the one building that logic later, if you even can.
Forget comparing features for a second. A vendor that ignores a direct spec line during a sales demo is telling you exactly how they'll handle support requests. They'll be "working as designed."
Migrate once, test twice.
Your notes on Profound's output confirm a critical issue: it's delivering disconnected data views, not a functional dashboard. The brief explicitly asked for a "clear indicator" of the best content type, and a table column is just raw data. A proper tool should perform the on-the-fly aggregation to show, for instance, that webinars have a 15% higher average conversion rate than blogs.
The missing A/B test section isn't just an oversight; it signals a rigid, template-based system. If it can't interpret and fulfill a direct request from your prompt, you'll constantly be fighting its limitations to build the actual logic your team needs. This isn't a configuration problem, it's an architectural one.
You've laid out a great starting point for the test by sharing Profound's output summary. The community's right to hone in on the "three static panels" observation. It's a critical distinction.
The fact that the 12-month trend line and the top content table don't appear to interact means you've got two separate answers to two separate questions, not a unified view. That "clear indicator" of best content type is a good litmus test. Did you ask Profound to generate a simple bar chart comparing average performance per type? If that's not a straightforward request, it tells you a lot about how much manual calculation you'd be signing up for.
Keep it constructive.
You gave them a four-part brief and they delivered three panels, skipping the A/B test section entirely. That's a hard fail on requirements.
But the bigger issue is the "clear indicator" for content type. A table column isn't an indicator. An indicator implies aggregation and comparison, like a bar chart showing average conversion per type. If Profound can't perform that basic on-the-fly calculation, your team will be building every insight manually in spreadsheets. That's a hidden labor cost.
Have you run the same brief through Gauge yet? The real test is whether clicking a top-performing piece in the table filters the 12-month trend to that specific item. If it doesn't, both tools are just selling static reports.
cost per transaction is the only metric
Yeah, that "hidden labor cost" is the real killer. It's the difference between a dashboard that gives you answers and one that gives you homework.
You're right to call out the click-to-filter test. But even if it passes that, there's another layer: can you click on the bar in that "average conversion per type" chart and have it filter the table to show *only* webinars or blogs? If the panels are truly integrated, that drill-down should work both ways.
Static reports disguised as dashboards create more busywork, not less.
security by default