Most BI waits for you to go looking. Proactive analytics watches your data and tells you what changed, why, and whether you are on track.
Here's a pattern almost every team falls into. Someone builds a beautiful dashboard. It goes live, everyone admires it for a week, and then it slowly fades into the background. Two months later a number is quietly going sideways and nobody notices, because noticing required someone to remember to open the dashboard, scroll to the right chart, compare it to last period in their head, and decide whether the movement mattered.
That's the core flaw of traditional business intelligence: it is pull-based. The insight sits there patiently, waiting for a human to come pull it out. If you don't go looking, you don't find it. And people are busy, so mostly they don't go looking until something is already on fire.
Think about how this plays out in practice:
- Dashboards nobody opens. The effort went into building them, not into making sure anyone reads them on the day a number breaks.
- Problems found late. A refund spike, a stalled signup funnel, a region that quietly slipped, all discovered in the monthly review instead of the morning it started.
- The interpretation tax. Even when you do open the dashboard, you still have to do the work: is this up or down versus last period, is it a real shift or just noise, and what actually caused it?
None of this is a failure of the dashboard. A dashboard is a great way to answer a question you already know to ask. The problem is that the most expensive moments in a business are the ones you *didn't* know to ask about. For the bigger picture on what BI is supposed to do, our complete guide to business intelligence is a good companion read. This post is about the half that pull-based BI leaves on the table.
Proactive analytics inverts the relationship. Instead of you going to the data, the data comes to you. The system continuously watches your latest numbers and surfaces three things, in plain English:
1. What changed — a metric moved meaningfully versus the prior period.
2. Why it changed — which segment or category drove the movement.
3. Whether you're on track — how the current trajectory compares to a goal you set.
The difference is the difference between a smoke detector and walking through the house every hour sniffing for smoke. Both can catch a fire. Only one of them works when you're asleep, distracted, or simply trusting that things are fine.
Concretely, "proactive" in VedaX Studio shows up in three connected places:
- The Insights feed — a running list of findings VedaX generates from your data: trends, drivers, goals, and anomalies.
- Anomaly scans — a scheduled scan that flags unusual spikes and drops you didn't ask about.
- The Insight Digest — an opt-in email that delivers those findings to your inbox on a cadence you choose, so the system reaches you even on the days you never open the app.
Notice what's missing from that list: a request. You don't query for any of it. The findings arrive. This is the same shift that makes automated reports and alerts valuable, taken one step further: instead of automating the delivery of a chart, VedaX automates the *judgment* of what's worth your attention in the first place.
The Insights feed is the heart of proactive analytics in VedaX. Each card is a single finding, written as a sentence a human would say, with two actions attached: Ask Buddy (open it up in conversation) and View data (jump straight to the underlying rows). There are four kinds of finding, and they map neatly onto the four questions you'd otherwise be asking yourself.
Trends — "a number moved." A trend finding tells you a metric changed by some amount versus the prior period. Revenue down week over week, signups up versus last month, average order value drifting. It's the headline: something is different now than it was before.
Drivers and contribution — "here's what moved it." A trend on its own raises the question "caused by what?" Driver findings answer it by breaking the change down into the segments behind it. Imagine a store whose total orders jumped, and the finding reads "records up, driven mainly by Completed." That's contribution analysis: attributing a top-line move to the parts underneath it. It's the most useful and most underused move in analytics, and we give it a full walkthrough in what drove the change: a plain-English guide to contribution analysis.
Goals — "are we on track?" Set a target and VedaX tells you whether your current pace lands you there. On track or off track, surfaced before the deadline instead of in the post-mortem.
Anomalies — "this is weird." Trends are about expected movement; anomalies are about the unexpected kind. An anomaly finding flags a spike or drop that breaks the normal pattern of a metric, the stuff you'd never have thought to put on a dashboard. The hard part is always telling a real anomaly apart from ordinary noise, which is exactly what trend, anomaly, or noise: how to read a moving metric is about.
The quiet hero here is that every card is a complete thought. You're not handed a chart and asked to interpret it. You're handed the interpretation, with the evidence one click away.
A feed that surfaces everything is just a dashboard with extra steps. The value of proactive analytics depends entirely on relevance: it has to show you the few findings that matter and stay quiet about the rest. Two things make that possible in VedaX.
First, it knows what your columns mean. Before VedaX can tell you a metric moved, it has to know which columns are metrics in the first place. It classifies every column by its role — metric, dimension, identifier, or date — so it analyses the right things and, just as importantly, doesn't analyse the wrong ones. It won't sum a customer ID or trend a postal code, because it understands those aren't quantities to be measured. This semantic layer is what we call the business dictionary, and it's worth understanding on its own terms: see metric vs. dimension: the two words that make analytics click for the concept, and the business dictionary and semantic layer for how VedaX builds and uses it.
Crucially, you get the final say. If VedaX mislabels something, you can correct it — "this isn't a metric," or "track this instead" — and your correction wins everywhere downstream. Good column roles are the difference between findings that read like a sharp analyst and findings that read like a confused spreadsheet.
Second, it ranks. From your latest data, VedaX assembles candidate findings — every meaningful trend, driver, goal status, and anomaly it can detect — and then prioritizes. A large move on a metric you care about outranks a tiny wobble on a minor one. The point of the ranking is to respect your attention: the feed is a short list of what changed and what's unusual, not an exhaustive audit. When a finding does land, you can immediately pull the thread with Ask Buddy, where the AI assistant lets you ask follow-up questions in plain English and get the chart or table back. That's the bridge from "here's what changed" to "so what do we do," which is the whole game of turning dashboards into decisions.
The Insights feed solves the "go looking" problem only if you open the feed. For the days you don't, there's the Insight Digest — an opt-in email that brings your insights to you.
It's deliberately opt-in and off by default, because nobody should be subscribed to an inbox they didn't ask for. When you do turn it on, you choose:
- Who receives it — the recipients you pick, not a blast to the whole org.
- How often — daily or weekly, depending on how fast your numbers move.
- Exactly when — a specific day and time, in your workspace's timezone, so it lands before your standup instead of at 3 a.m.
There's also an org-wide admin view for Insight Digests, so whoever runs analytics can see and manage the digests across the workspace rather than each person guessing in the dark.
The digest is what turns proactive analytics into a habit instead of a feature. A weekly digest landing every Monday at 8 a.m. is, functionally, a founder's weekly metrics ritual that assembles itself. You don't have to remember to pull the numbers, and you don't have to remember to read the dashboard, because the short list of what changed is already sitting at the top of your inbox. We go deep on cadence, recipients, and what makes a digest worth opening in the analytics email digest: put your key numbers on autopilot.
You don't need a data team or a six-week rollout. The shift from pull to push is mostly about turning a few things on in the right order and trusting the feed to earn your attention.
1. Connect the source that actually drives decisions. Start with one — your database, your store, your CRM, a Google Sheet, whatever holds the numbers you'd be upset to find broken late. Connections in VedaX are read-only, so you're exposing the system to your data, never the other way around; it never writes back.
2. Fix the column roles before you trust the findings. Open Data Explorer, glance at how VedaX has classified your columns, and correct anything off. Two minutes here pays for itself: a metric mislabeled as an identifier is a finding that never fires, and an identifier mislabeled as a metric is a finding that's nonsense. Your corrections stick everywhere.
3. Read the Insights feed for a few days before automating anything. Let it surface trends, drivers, goals, and anomalies. Use Ask Buddy when a card raises a question and View data when you want to see the rows. You're calibrating: learning which findings are signal for your business and getting a feel for the ranking.
4. Set the goals that matter. A goal turns a vague "how are we doing" into a crisp on-track / off-track finding. Even two or three well-chosen targets change the feed from descriptive to accountable.
5. Turn on the digest once the feed feels trustworthy. Pick daily or weekly, pick the recipients, and pick a time that fits an existing rhythm — before a standup, ahead of a Monday review. Now the system reaches you whether or not you open the app.
6. Let anomaly detection run in the background. The AI-driven anomaly detection scan is the part that catches what you never thought to monitor. It's scheduled, so once it's on it just works, flagging the unusual movements that would otherwise surface weeks later in a review.
Done in that order, proactive analytics stops being a thing you configure and becomes a thing that quietly watches your back. The dashboards don't go away — they're still the right tool for the questions you already know to ask. What changes is that you stop relying on your own discipline to catch the questions you didn't. The insights find you.
A dashboard is pull-based: it waits for you to open it and interpret the numbers yourself. Proactive analytics is push-based: the system watches your latest data and surfaces what changed, why it changed, and whether you're on track, then delivers those findings to a feed and, optionally, to your inbox. You don't have to remember to go looking.
No. VedaX assembles candidate findings from your data and then ranks them, so the feed is a short list of what actually matters rather than an exhaustive audit. A large move on a metric you care about outranks a tiny wobble on a minor one. The Insight Digest is also opt-in and off by default, and you choose its cadence and recipients.
It classifies every column by its semantic role — metric, dimension, identifier, or date — so it measures the things that are actually quantities and ignores the ones that aren't. That's why it won't sum a customer ID. If it mislabels a column, you can correct it, and your correction applies everywhere downstream.
They're connected but distinct. The Insights feed is the in-app list of findings you can browse anytime, with Ask Buddy and View data on each card. The Insight Digest is an opt-in email that delivers those insights to chosen recipients on a daily or weekly cadence, at an exact day and time in your workspace's timezone, so they reach you even when you don't open the app.
VedaX connections are read-only. The platform reads your data to analyse it and never writes back to your source, whether that's a SQL database, a store like Shopify or Stripe, a CRM, or a spreadsheet. You can start with a single source and expand once you trust the findings.
See it in VedaX: AI business intelligence for small businesses