The challenge

Originally focused on partner analysis, the agency’s IQBlade platform has now been expanded to include analysis of any UK company, with the addition of many new data sources and extra functionality.

Further expansion is planned, so the company is looking for ways to automate the classification of organisations within the ecosystem, as Ben Abraham explains:

“IQBlade operates in a very specific segment of the technology channel. This encompasses several different business types, all of whom sell, implement, and manage technology solutions for their end-user clients, on behalf of vendors such as Amazon Web Services, Microsoft, IBM, Lenovo, Cisco etc.

“The IQBlade platform has a classification model of 14 different company types in the technology channel. Many of the companies had been manually classified over time and we saw an opportunity to automate this process to drive more scale to the platform, rather than relying on human classification.”

Benefits of outsourcing to PhD student

Although IQBlade has an in-house team of data scientists and engineers, it decided to outsource the classification project to a PhD student at the Liverpool Centre for Doctoral Training for Innovation in Data Intensive Science (LIV.INNO).

PhD students from LIV.INNO support businesses with bespoke projects, offering specialist skills in prediction, risk analysis, visualisation, and scenario modelling.

Ben explains their rationale: “We wanted a fresh perspective to try to solve this issue and felt that bringing in somebody who was external to the team might approach the challenge without the bias created by knowledge of the subject matter.

“We had previously collaborated with LIV.INNO and were very pleasantly surprised by just how advanced the skills and capabilities of the team and the cohorts of students were.

“We also realised that many of the problems that we were looking to solve had similarity to solutions in other fields of study.

“We felt that it would be interesting to see if some of that knowledge could be re-applied to our use case.”

Building capacity for IQBlade

Data Scientist and LIV.INNO PhD student Alberto Acuto was engaged by IQBlade for the industry placement. He saw it as an exciting opportunity to apply his skills in signal processing to a new challenge.

He explains: “The project focussed on building classification models to identify tech companies. It used text descriptions and economic data gathered by the IQBlade database.

“I developed software to extract (requests), process raw text data (natural language toolkit, pandas) and analyse (clustering, statistical analysis) information from several company categories to create a prediction model. Further tools were created to assess the accuracy of the predictions using cross-validations methods between the models and by adding economic constraints.”

Ben was pleased with his approach: “Alberto needed very little supervision. We just spent some time at the start of the project to explain the challenge, gave access to our datasets and then shared what we had already tried beforehand to give some context and guidance.”

Outcomes and conclusion

“The ideal scenario would have been to build a single model that was able to classify all 14 types of partner – but we recognised that that was unrealistic,” continues Ben.

“Having said that, Alberto was able to build a model that accurately classified the three main partner types (which between them account for almost 60% of the technology channel), so that has helped to accelerate the rate at which we can ingress new data into the platform.”

“I would certainly recommend other companies work with LIV.INNO on projects in this field, or in any area where there are challenges which have a data theme. I’m sure there are many businesses in the region who would improve their commercial performance through this program.”

IQBlade offers digital marketing services based on a data-led approach to channel development. By harnessing vast quantities of data and leveraging its platform technology, the agency enables tech vendors and their ecosystem partners to maximise sales growth.

IQBlade has developed over 150 campaigns for some of the world’s largest technology clients, featuring video, brochure, social ad, landing page and email creation.

Find out more about IQBlade at iqblade.com

The Liverpool Centre for Doctoral Training for Innovation in Data Intensive Science, provides a hub for training students in managing, analysing and interpreting large, complex datasets and high rates of data flow.

The Centre features a unique training approach, addressing some of the biggest challenges in data intensive science to tackle a growing skills gap in this important area. This includes cutting edge research projects, and a targeted academic training programme, complemented by placements in industry tackling real world challenges.

Find out more at livinno.org.