AI Powered Campaign Intelligence for Multi Channel Marketing

Deployment

Tom Lorimer

Chief Executive Officer

The problem

‍

How can you bring campaign performance data from multiple platforms into one place? Thereby, making it easier to understand results, compare performance and answer questions without manually pulling together reports? This was the challenge we explored within‍, a UK B2B media, events and data group with more than 400 employees.

‍

Its marketing solutions team runs integrated campaigns across display advertising, sponsored content, newsletters, lead generation, webinars and events. However, each channel reports through a different system, with no shared client ID connecting the data. As a result, consolidated campaign reports had to be put together manually for a small number of key accounts, while performance benchmarks were recalculated in Excel twice a year.

‍

01

Fragmented data

Campaign performance was spread across multiple platforms, making it hard to see a complete picture of client results.

02

Manual reporting

Bringing together data from different channels was time-consuming and difficult to scale.

03

Limited benchmarking

Comparing performance across accounts meant manual calculations and infrequent updates.

04

Difficult data access

Straightforward questions often meant navigating several systems, or waiting for someone to pull the numbers together.

‍

What we did

‍

Following a four week discovery phase, we developed an deployed an AI assisted campaign intelligence platform, using input from William Reed's product and commercial teams. The prototype brings together campaign data from seven source systems into a single dashboard, giving users a consolidated view of performance across channels and marketing funnel stages.

‍

01

Unified reporting

One dashboard showing campaign performance across channels and funnel stages.

02

AI-powered summaries

Automatically generated insights, with every figure checked against the underlying data before it's displayed.

03

Natural language queries

Ask about campaign performance in plain English, translated into database queries.

04

Peer benchmarking

Compare campaign performance against other accounts and relevant benchmarks.

05

Lead and event forms

A simple way to capture relevant campaign interactions.

‍

It also incorporates safeguards around data accuracy and privacy.

‍

Ambiguous client identity matches are flagged for human review rather than automatically guessed.
Personal data is kept separate from the language model
AI generated figures are validated before appearing in summaries

‍

The impact

‍

The prototype has the potential to automate an estimated 80 to 160 person days of manual reporting each year, across 20 key accounts currently reported on quarterly. It also provides a foundation for more accessible campaign intelligence, allowing teams to explore performance, identify patterns and compare results without relying on manually assembled reports.

‍

< back to Projects

< previous

Next >