Reporting that explains why the number moved, not just that it did

Reporting that explains why a number moved attributes a change to its cause. A unified marketing and sales dashboard brings every channel, the CRM and finance into one reconciled view, then uses AI to read each movement and name what drove it: the campaign, segment or deal stage that shifted, and by how much. A plain dashboard tells you the figure changed. This kind tells you the figure changed and why, in the same place, so the meeting starts from the answer instead of the question.

What is a unified marketing and sales dashboard?

A unified marketing and sales dashboard is one reporting view that reads from a single reconciled layer of data, so marketing, sales and finance see the same figures rather than three versions that disagree. It pulls campaign results, pipeline events and revenue into one place and presents them as a connected story: spend leads to leads, leads to opportunities, opportunities to closed revenue.

The word that earns its keep is unified. A dashboard that simply embeds charts from your ad platform, your CRM and your finance tool side by side is not unified. It inherits every disagreement underneath it. A genuine unified dashboard sits on top of reconciled data, which is why it has to follow the work of unifying marketing and sales data with AI rather than precede it.

Why do most dashboards show what changed but not why?

Most reporting is built to display metrics, not to interpret them. The dashboard tells you pipeline fell 18 percent this month. It does not tell you that the fall sits entirely in one region, or that a single paused campaign accounts for most of it. So the figure triggers an investigation rather than a decision, and someone spends an afternoon opening five other tools to reconstruct what the dashboard could have said in a sentence.

The deeper cause is the data underneath. Research on marketing silos compiled by Amra and Elma in 2025 found that 47 percent of marketers name data silos as the single biggest barrier to getting useful insight from their data, and only about half of organisations report a genuine single source of truth for sales and marketing. When the numbers in a dashboard are drawn from systems that do not agree, the dashboard cannot explain a movement, because it cannot trace the movement back to a clean record. It can only show the surface.

How does AI explain why a number moved?

Once the data is reconciled, explaining a change becomes a job AI can do well. A model can read the current figures against the prior period, decompose the difference into its parts, and rank the drivers by how much each contributed. Pipeline is down 18 percent: 12 points from the North campaign pausing, 4 from a slower sales stage, 2 from seasonality. The model writes that in plain English next to the chart.

In a setup we would build, the moving parts are straightforward. The unified data layer holds the reconciled records. An AI layer, running a model such as Claude, reads the movement, attributes it and answers follow-up questions in the dashboard itself, so a manager can ask "why did conversion drop in the South?" and get a traced answer. Where the AI needs to reach a live system to check a detail, the Model Context Protocol, or MCP, gives it a standard way to connect, and the Claude Agent SDK lets you run the variance check on a schedule, so the explanation is waiting on Monday morning rather than requested after the fact. For UK firms, that processing can sit in a UK region such as AWS Bedrock UK South, which keeps customer records in the country and shortens the conversation with your security team.

What is the difference between live reporting and a monthly report?

A monthly report is a photograph of a moment that has already passed. By the time it is circulated, the business has moved on, and decisions taken from it are decisions about last month. Live reporting reads from the unified layer continuously, so the dashboard reflects the state of the business now.

The gap matters more as buyers speed up. Around 89 percent of B2B buyers now use AI search during the buying process, according to the Digital Agency Network generative engine optimisation statistics for 2026, and a typical B2B purchase involves a buying committee of six to ten people, on Gartner's 2023 research. When a committee that size is moving across channels in real time, a report that lags three weeks behind them is reporting on a different deal. Live reporting lets you react to a change while it is still in front of you, which is the entire commercial point of reporting in the first place.

What should a board level revenue dashboard show?

A board does not need forty metrics. It needs the few that describe the health of the revenue engine and, for each, the reason behind the current reading. Revenue against plan, pipeline coverage for the next two quarters, the cost of acquiring that pipeline, and net movement since the last meeting, each annotated with what drove it.

The annotation is what separates a board level dashboard from an operational one. "Pipeline coverage is 2.8 times, down from 3.4, driven mainly by a slower top of funnel in the industrial segment" is a sentence a board can act on. The raw number on its own invites ten minutes of questions the dashboard should already have answered. Revenue reporting at this level is less about more data and more about the data carrying its own explanation, which is exactly what AI on a reconciled layer makes affordable to produce every month without an analyst rebuilding the deck by hand. The same approach extends naturally into predictive analytics, where the dashboard moves from explaining the past to flagging the likely next move.

How do role based dashboards work?

Role based dashboards solve a simple problem: the same reconciled data needs to look different depending on who reads it. Marketing wants channel and campaign detail. Sales wants account and pipeline movement. Finance wants attributed revenue and cost. The board wants the four numbers above. One source, several views.

Because every view draws from the same reconciled layer, the figures reconcile across roles even though the framing changes. Marketing's campaign that created pipeline and finance's booked revenue are the same money seen from two ends, so the argument about whose number is right does not start. This is the practical payoff of doing the data work first, and it is the same foundation that lets you push intelligence back into day to day tools through AI integration with CRM, ERP and ecommerce, so the people who act on a number see it where they already work.

Where should you start?

Start with the question your leadership keeps asking after the fact: why did that number move? Trace it to the systems that should answer it, reconcile those first, and build the dashboard on the clean layer. You do not need every source connected on day one. You need the two or three that hold the records your most important figures depend on.

If you are sequencing a wider AI programme, reporting sits just after the data foundation, because a dashboard is only as honest as the data beneath it. Our practical 2026 framework for implementing AI in a B2B business sets out where reporting falls in the order of work, and the dedicated AI marketing and sales unified reporting dashboards programme page covers how Teylu builds and runs it. Get this right and the weekly review changes character: it opens with the reason, not the riddle.

FAQ

How do you build a unified marketing and sales dashboard?
You start with one reconciled layer of data that joins your marketing channels, CRM and finance system, so a lead, an account and a deal mean the same thing everywhere. You then build the dashboard on top of that layer rather than on the raw tools, and add an AI layer that reads each movement and explains what drove it. The order matters: reconcile the data first, report second.

What does it mean for a dashboard to explain why a number moved?
It means the dashboard attributes a change to its cause, not just records that the change happened. When pipeline drops, the dashboard names the segment, campaign or deal stage behind the drop and quantifies each contribution, so the reader sees the figure and the reason together instead of opening five other tools to work it out.

What is the difference between live reporting and a monthly report?
A monthly report is a snapshot that is already weeks old by the time it is read, so decisions are made on a past state of the business. Live reporting reads from the unified data layer continuously, so the dashboard reflects current truth. The practical gain is reacting to a change while it is still happening rather than reviewing it after the quarter has closed.

Do you need a data warehouse to build a unified marketing and sales dashboard?
Not to begin with. Many mid-sized B2B firms reach a usable unified dashboard by connecting their CRM, ad platforms and finance system through a unification layer and building reporting on that, with AI handling the matching underneath. A warehouse helps once data volume grows, but you can start without one and add it later if scale demands it.

Is AI reporting safe under UK and EU data rules?
It can be, with the right setup. UK data protection sits under the ICO, and the AI processing can run in a UK region such as AWS Bedrock UK South so records stay in the country. The EU AI Act adds governance duties, with full high-risk enforcement from 2 August 2026 and penalties up to 35 million euro or 7 percent of global turnover. Reporting is usually lower risk, but you still need an AI inventory, a named owner and documentation.