Overview: Knowledge graphs in Practice
This article from Revology Analytics explains knowledge graphs in the context of modern pricing analytics and revenue growth management. It draws on real engagements with mid-market and enterprise clients to turn knowledge graphs from a buzzword into a measurable commercial capability. Read on for the full perspective, and see our related reading for additional depth.
Table of Contents
In our recent webinar titled “Supercharge Your Revenue Growth Strategy with Knowledge Graphs,” we discussed how this unique commercial analytics approach makes understanding multi-dimensional business data more accessible and substantially faster, helping you make smarter decisions and drive Profitable Growth quicker.
Knowledge Graphs are a highly efficient way to represent your otherwise disconnected sales, marketing, customer, competitor, and supply chain data to derive holistic, integrated, and actionable Revenue Growth Analytics insights in minutes vs. days.
Although Knowledge Graphs, like those built with Neo4j, have been around for a while now, outside of core areas like search engines, social networks, and fraud detection, they are not yet prevalent for Revenue Analytics problems. It’s an emerging field in Sales & Marketing Analytics (think customer journeys, next-best-sales-action, multi-touch attribution) but sparsely used for other areas like Pricing & Promotion Optimization, Customer Analytics, and Sales Optimization.
Knowledge Graphs naturally excel at storing, visualizing, and analyzing multi-dimensional relationships. They can be compelling analytical assets for any enterprise that has a relatively high purchase frequency and that can track customer touchpoints and engagements somewhere (CRM, marketing data platforms, etc.).
What would typically take layers of nested queries and days or weeks of analyses can be accomplished using purpose-built Knowledge Graphs tailored to the particular industry and company.
Here are some contrasts between traditional descriptive/diagnostic analytics vs. doing a simple query with Knowledge Graphs:
Retail:
How does in-store staff training influence store sales of Product XYZ within 90 days of training?
Traditional Analytics: You must merge in-store sales from your transactional database with training records and perform some descriptive pre- vs. post-analysis or run a regression / ML model. A multi-step process typically takes a day for a good analytics team to answer.
Knowledge Graph: a single query can trace the relationship between specific training sessions and subsequent in-store sales of Product XYZ. You can also apply a Machine Learning model quite easily to quantify the impact.
See how your pricing compares
Our 2025 report looks at how more than 150 business leaders handle pricing and promotions. See what the best companies do differently, and where most fall short.
Consumer Products:
What is the typical sales decline when Competitor B launches a product in Category ABC? How does our sales performance change if we counter price promotions with a 10%, 20%, or 30% discount?
Traditional Analytics: You’d have first to identify competitor product launches from syndicated data, then pull and analyze your sales data during those periods, bifurcated by promotional vs. baseline periods. It’s time-consuming and might not account for other influencing factors. Again, a 1-2 day turnaround for most analytics teams.
Knowledge Graphs: since we can easily map out the relationship between Competitor B’s product launches, our sales trends, and our promotional activities (all represented in our graph), we can quickly answer this question.
Of course, the emphasis is building a Knowledge Graph that’s “purpose-built” for the particular industry, company, and even domain (i.e., Sales, Marketing, Supply Chain, etc.). Careful design and co-creating with internal stakeholders are paramount as that will solely determine the analytical questions you can quickly answer.
The recording is available here for those who missed the live session or wished to revisit the insights.
If you have trouble viewing the below presentation, you can also download it here.
Frequently asked questions about knowledge graphs
What is a knowledge graph in revenue growth analytics?
A knowledge graph is a highly efficient way to represent otherwise disconnected sales, marketing, customer, competitor, and supply chain data. Graphs excel at storing, visualizing, and analyzing multi-dimensional relationships, so teams can derive holistic, integrated, and actionable revenue growth analytics insights in minutes instead of days.
Where are knowledge graphs used today?
Knowledge graphs, such as those built with Neo4j, have been around for a while in core areas like search engines, social networks, and fraud detection. In sales and marketing analytics they are an emerging field, for example in customer journeys, next-best-sales-action, and multi-touch attribution, but they are still sparsely used for pricing and promotion optimization, customer analytics, and sales optimization.
Which companies benefit most from knowledge graphs?
Any enterprise with a relatively high purchase frequency that tracks customer touchpoints and engagements somewhere, such as a CRM or a marketing data platform, can turn a knowledge graph into a compelling analytical asset. For these companies, a purpose-built graph can accomplish what would otherwise take layers of nested queries and days or weeks of analysis.
How do knowledge graphs compare with traditional analytics?
Take the retail question of how in-store staff training affects sales of a product within 90 days. Traditional analytics merges transactional sales with training records and runs a pre- versus post-analysis or a regression or ML model, a multi-step process that typically takes a good team a day. A knowledge graph traces the link between training sessions and later sales in a single query.
How can a consumer products company use a knowledge graph?
It can find the typical sales decline when a competitor launches a product in a category, and how sales change if it counters with 10%, 20%, or 30% promotional discounts. Traditional analysis of syndicated and internal sales data takes most teams 1 to 2 days and may miss other factors, while a graph that maps launches, sales trends, and promotions answers the question quickly.
How do you build an effective knowledge graph?
Make it purpose-built for your industry, your company, and even the domain, such as sales, marketing, or supply chain. Careful design and co-creation with internal stakeholders are paramount, because the design alone determines which analytical questions the graph can answer quickly.
Related Reading
- Dynamic Pricing for B2B: Real‑Time Strategies to Optimize Wholesale and Distribution Margins
- RA Quick Insights Video Series: Driving rapid margin actions with transactional data analysis (Part II. – Discount Curve Analysis)
For broader industry perspective on pricing analytics and revenue growth management, see McKinsey’s Growth, Marketing & Sales insights.
Want a second opinion on your pricing?
Tell us what’s going on with your pricing. One of our partners will get back to you the same day or the next.
