Transcript
Nobody wants another dashboard
23 May 2026
19 min
The Friday Afternoon Report
So picture this scenario. It's Friday afternoon.
The best time of the week, right?
And today, specifically, let's say it's May 22nd, 2026. You are just staring at your screen, running out the clock until the weekend, and ping, the weekly performance report hits your inbox.
Oh, the dreaded weekly report.
Exactly. So you open it up and you see that overall conversion is down like 2.3 percent, average order value is up slightly, the return rate is stable. So you do that thing that we all do, scan it.
Yeah, you scan the numbers, you mentally flag a couple of metrics to bring up in the Monday morning standup, and you close the tab.
But here's the thing. By the time Monday actually rolls around, those numbers are already completely stale.
Entirely stale. And more importantly, that report never actually told you what to do about any of those numbers. It just gave you a chore for next week.
Yeah, it's just a snapshot of the past that offers frankly absolutely no map for the future. You are essentially looking at a digital rearview mirror and trying to use it to steer the car.
Which perfectly sets up our mission for this deep dive today. We're analyzing a really thought-provoking article from lightsout.ai and it's titled Nobody Wants Another Dashboard.
It's a great piece, it really is. So our goal is to explore why our current corporate obsession with data visualization, all those beautiful colorful charts and graphs that we just stare at on a daily basis, is actually slowing down our decision-making.
Slowing it down drastically, right.
The Decision Shortage
And we're going to look at what the future of interacting with data is structurally supposed to look like. To really set the stage for why this shift matters so much to your daily workflow, we need to understand the core premise of the article. The fundamental argument here is that modern businesses do not have a data shortage.
Not at all. I mean, you have more data than you could ever possibly process. What businesses actually have is a decision shortage. We are just drowning in information but we are completely starved for action.
Okay, let's unpack this. Because when we talk about drowning in information, we are definitely talking about the dashboards. The usual suspects.
Oh yeah, we're talking Looker, Metabase, Power BI, Tableau. Maybe you even have all of them running at your company right now because you inherited them from different teams over the years.
That happens all the time. There was always someone who needed just one more view for specific projects, so they commissioned a completely new dashboard.
And to be fair, these tools do exactly what they promised to do. They take your data, they put it into charts, and they show you exactly what happened.
But the article argues that knowing what happened is just table stakes now. It is the absolute bare minimum.
Exactly. And the author introduces a concept that I think perfectly captures the futility of this approach. They call it data archaeology.
Data Archaeology
Data archaeology. I love that, it's so accurate.
Think about the environment most of us operate in, particularly in something like a VC-backed e-commerce company. Super fast-paced. In that world, inventory shifts daily, new marketing campaigns launch constantly, customer behavior changes with the weather or a viral social media trend or some macroeconomic blip. So in an environment that fast and that volatile, looking at last week's aggregated numbers isn't looking at actionable data. It's archaeology. You are literally digging up ruins to guess how a civilization died rather than trying to save the people who are currently living in it.
That makes a lot of sense. Especially the frustration of brushing the dust off a drop in conversion from last Tuesday. It's just exhausting to constantly be looking backward.
Yeah, it's useful for long-term pattern recognition, sure. But it is entirely useless for deciding what your team needs to do right now, in this very minute, to fix a live operational problem.
Okay, well let me push back on this a bit though. Because it feels like driving a car where the check engine light comes on. The dashboard in my car doesn't tell me how to fix the engine, it just alerts me that something's wrong. The manual essentially just says hey, the engine's definitely doing something weird, take it to a mechanic. So isn't the whole point of a business dashboard just to be that alert system? Is this really a failure of the software tools, or is this just a lazy team problem? Shouldn't a smart, well-paid team know how to look at that flashing light and just go dig into the data to find the root cause themselves?
That's a fair question. But it assumes the digging is a quick, frictionless process, which completely ignores the realities of modern data architecture.
Fair.
The issue isn't laziness at all. The issue is that dashboards tell you what is wrong, but their underlying structure prevents them from telling you why and what to do about it. Because they're just a surface-level view.
The Life Cycle of a Dashboard
Exactly. Think about the life cycle of a typical dashboard. It gets built with the best intentions, it gets presented in a big flashy kickoff meeting, everyone claps.
Everyone claps, you bookmark it.
Then gradually over a few weeks it gets forgotten. And it gets forgotten because staring at a pie chart doesn't actually help anyone do their job faster. A graph showing a conversion decline answers the first question: is there a problem? But it totally abandons you when you ask the next two critical questions, which are: why is this happening, and how do we fix it? Because looking at the chart is passive. It's just a static image of a problem, and the entire burden of translation falls on the user.
Right. And this massive void between measuring a metric and taking action to improve that metric has a real, quantifiable cost. It costs human hours, it causes decision fatigue, and it leads to organizational paralysis.
Which is huge.
Huge. And the article gives a concrete example of what this looks like in practice, and it perfectly illustrates the human cost of this gap.
The 45-Minute Scavenger Hunt
Yeah, I definitely want to walk through this scenario. And I want to put you, the listener, directly in the hot seat for this. Let's track the exact steps you have to take on a Tuesday morning when your CEO Slacks you the dreaded Slack message.
Oh yeah, they ask why revenue in a specific product category is suddenly sitting below forecast. Your dashboard flashed that warning light, conversion dropped. So now what?
The scavenger hunt begins. The 45-minute scavenger hunt. Let's meticulously track this exhausting workflow just to figure out what the dashboard is trying to tell you.
So step one: you open your commerce platform, maybe Shopify or Salesforce. You're checking the pricing, making sure no automated discounts went rogue.
Okay, it looks fine. Step two: you leave that system entirely and you have to open the PIM, the product information management system.
Yeah, because you're verifying that the product descriptions are actually complete, that the images aren't broken, and that the language localizations are actually displaying properly.
Exactly. But I still don't have the answer. Everything looks fine on the surface, so I have to keep digging. Step three: you open your analytics tool, like Google Analytics or Mixpanel, to see if there was a sudden drop in traffic to those specific product pages.
And then step four, you jump into the marketing platform. You have to check if the ad campaigns driving traffic to that category are actually running, or if maybe a budget capped out unexpectedly.
And then step five: you log into the inventory system or ERP just to confirm that the products are physically in stock in the warehouse and available to ship.
So after 45 minutes of logging in and out of different software silos, comparing tabs on my monitor, and just trying to hold these fragmented pieces of data in my head, I finally discover the culprit.
And what is it?
Three products in that specific category have had incorrect prices on the German market since Wednesday. Because of a silent sync error between the ERP, the enterprise resource planning software, and the commerce platform.
What's fascinating here is that nobody caught this massive error when it happened, because absolutely no single system was looking at the full picture.
No one had all the pieces. Your dashboard relies on a data warehouse. The data warehouses are usually updated in batches, maybe overnight. So the data is aggregated, and aggregation just strips out the granular anomalies.
So it just smooths everything out.
Exactly. A three percent drop in overall conversion completely hides the fact that three specific products in Germany are priced at zero dollars. The dashboard did its job technically. It told you the aggregated number looked bad. But finding the granular cause and fixing it cost you half a day of productivity.
Now multiply that single 45-minute scavenger hunt by every anomaly, every weird metric, every single week, across every single team in your entire company.
That is the crippling hidden cost of reporting without action.
From Dashboard to Command Center
It's staggering when you frame it like that. If the real pain point isn't a lack of data but the sheer amount of time we spend trying to translate a colorful graph into an actual human action, then buying a prettier dashboard software isn't going to fix anything.
Not at all. It requires a total structural shift in how our software actually works and connects. Which brings us to the core thesis of the proposed solution. If we abandon the dashboard, the article proposes moving to what they call a command center.
Okay, let's redefine the workspace then. How does a command center actually function differently? If I'm logging in on a Tuesday morning, what does my screen look like compared to a standard dashboard?
Well, the difference is structural, really coming down to reporting versus working. Dashboards are built exclusively for reporting. They're something you look at. A command center is built for working. It is something you work in. Structurally it doesn't just pull from a stale data warehouse. It relies on a unified operational data layer.
What does that mean exactly?
It means it sits across the live APIs of all those multiple systems we just talked about. The PIM, the ERP, the analytics. It cross-references operational logic in real time, identifies deviations, and crucially, it connects every single deviation to a suggested action.
The German Pricing Error, Revisited
Okay, here's where it gets really interesting. Let's go back to the scenario with the German pricing error. If I have a dashboard, the dashboard just says conversion dropped three percent last week, end of sentence, good luck.
Go hunt.
But if I have a command center, how does that experience change?
The command center says: product X has the wrong price on market DE since Wednesday. Here is the exact price discrepancy. Here is the revenue that has been affected so far. Here is the code correction to fix the sync error. Click here to approve or modify this fix.
Wait, let me stop you there. If the system is so smart that it can identify the missing description, spot the sync error, and literally write the code correction, why does it even need a human to click approve? Why doesn't it just fix it automatically?
Yeah, that is the crucial line between a helpful tool and a dangerous liability.
Okay.
Full automation without human oversight in complex enterprise environments usually ends in disaster.
I can see that.
What if that price drop wasn't a sync error but a strategic flash sale initiated by a regional manager that the central system simply hadn't tagged correctly yet?
Oh, and if the system auto-corrects it, you just ruined a marketing campaign.
Thanks. So the command center provides governance. It handles the heavy lifting of the diagnostic scavenger hunt but it leaves the final strategic context, the human context, to you. And over time, as you approve these fixes, you are actually training the model on your company's specific operational logic.
That makes perfect sense. It acts as the ultimate analyst, but I am still the decision-maker.
Marketing and Live Customer Data
So how does this shift actually reorganize the workflow of different departments? Let's take the marketing team for example. Does this operational data layer work for customer data too? Say I'm trying to run a campaign today.
Oh, it completely transforms marketing workflows. Think about how a dashboard presents customer data right now. It gives you a pie chart showing new versus returning customers. That is interesting but useless for launching a campaign today. The command center uses the operational data layer to perform live RFM analysis: recency, frequency, and monetary value.
Okay.
It doesn't just show you a chart. It segments customers by their actual up-to-the-minute behavior and surfaces their category affinities.
So instead of a chart, it essentially gives me a targeted list.
Exactly the opposite of a chart. It turns raw purchase data into a prioritized action list. It highlights who is ready to buy what right now.
Wow.
So the system might flag a segment of high-value customers who haven't purchased in 90 days, but they've been browsing the new winter catalog for the last 48 hours.
Specific.
Exactly. The command center generates the segment, suggests the optimal discount code based on past conversions, and presents it to the marketing manager as a pre-packaged campaign just waiting for deployment.
Gap Detection Done Right
See, the feature that really struck me as a game-changer is their approach to gap detection. Because usually a dashboard will give you an incredibly vague system health metric like 'data quality is at 73.' What does that even mean?
What on earth am I supposed to do with 73? Where is the 27 that's missing? I just have to go hunt for it. And because the dashboard is relying on aggregated batch data, it can only give you an aggregate score.
But the command center, sitting on that active operational layer, it moves completely past vague metrics.
It gives you a specific, actionable alert. It will tell you: product 4521 is missing a German description in the commerce platform and has a price mismatch between the ERP and Shopify. The level of context provided there eliminates the investigation phase entirely. It tells you what happened, when it started, which specific system is the source of the problem, and exactly what the recommended fix looks like.
The ultimate goal is that your teams go from investigating, like 'hey, something seems off with our numbers,' to confirming: 'here is the issue, here is the fix, I have approved it.'
That's a massive shift.
A New Monday Morning
It is.
When your operations team opens their workspace on a Monday morning, they never have to check six different tools just to understand the state of the business. They are greeted with a prioritized list of what needs their attention, armed with the exact context they need to act immediately. Deviations flagged, causes identified, fixes suggested.
And because it is a workspace and not a static report, all of these decisions are automatically logged in an audit trail.
Exactly. So when you have to do your monthly review, the report has essentially written itself based on the actions you actually took to fix the business, rather than just a summary of how the business broke.
The Paradox of Frictionless Decisions
If we pull back and look at the broader implications here, this completely redefines the core purpose of business technology. The ultimate goal of our software systems shouldn't just be to store information or visualize information. The goal must be to radically reduce the friction between discovering a problem and deploying a solution. Dashboards highlight the friction. Command centers eliminate it.
So what does this all mean for you? If you are listening to this and you are the one putting together those Friday afternoon reports, or maybe you are the executive receiving them, you really have to rethink your relationship with data.
You really do. You are smart. Your team is smart. You do not need more data. You certainly don't need a fancier pie chart. You need data that has already done the heavy lifting of cross-referencing your systems.
You need data that has been turned into decisions that are just sitting there waiting for your strategic approval.
Yes. The entire shift in modern business intelligence is moving away from generic reporting that merely describes the past and moving toward building customized command centers that actually facilitate the work of the present.
Exactly. Nobody wants another dashboard. They want to know what to do next.
Are We Outsourcing Our Own Expertise?
But if we connect this to the bigger picture, this shift raises a really profound philosophical question about the future of work. And it is one we really need to debate as we look toward this era of automated decision-making.
Oh, interesting. Where are you going with this?
Well, we are talking about eliminating friction. But if our data systems evolve perfectly into these seamless command centers, if the software merely presents us with a neatly curated list of decisions waiting for approval, what happens to the human skill of critical analysis?
Wow. That is a massive question. Are you suggesting that the friction we just spent 20 minutes complaining about actually serves a purpose?
In a way, yes. Think about the 45-minute scavenger hunt. It is exhausting, it's inefficient, and it's terrible for the company's bottom line. But in doing that digging, what is actually happening to the category manager? They are learning the intricate, messy realities of how their business actually operates. They are learning how the PIM talks to the ERP. They are building an intuitive mental map of the company's systems.
Wait, you mean the scavenger hunt is actually training us? It is forcing us to understand the plumbing of the business?
If the machine does all the diagnosing and the machine recommends the exact fix every single time, how does a junior employee ever develop the deep operational knowledge required to become a senior leader?
That's a great point. If we remove all the friction of investigation, do we eventually lose our intuitive understanding of our own businesses? The risk is that we become nothing more than human button-pushers, just blindly rubber-stamping the AI's choices without truly understanding the mechanics of why things are breaking in the first place.
That is a heavy realization. By outsourcing the diagnosis, we might actually be outsourcing our own expertise. Sure, we save 45 minutes today, but five years from now, does anyone in the company actually know how the whole machine works together?
Probably not. If the command center goes down, we wouldn't even know where to start looking. We would just be staring at a blank screen.
It creates a real paradox. We desperately need these tools to handle the sheer volume and velocity of modern data.
We cannot survive on dashboards.
Definitely not. But as we transition to command centers, leaders have to figure out how to artificially inject friction, or training, back into the system. Otherwise you aren't building a team of strategic commanders. You're just training a biological mechanism to click approve when the software prompts it.
That is an incredibly provocative thought to leave on. Are we building a world where we use the technology to elevate our strategy, or are we just becoming the final localized API endpoint for a machine that understands our business better than we do?
It's something to think about. It really is. Well, thank you so much for joining us on this deep dive. The next time that Friday afternoon report hits your inbox, take a hard look at it. Ask yourself: are you looking at a workspace that helps you take action, or are you just doing data archaeology?
Think about it. See you next time.
Want to discuss what you heard? We are always happy to talk.
Get in touch →