Digital Energy Journal- Calculating the ROI of Cognite software

Join | Existing Member Login

You are Home » News » View Article

Calculating the ROI of Cognite software

Tuesday, January 2, 2024

Forrester Research calculated the ROI of Cognite's 'data contextualisation' software to oil and gas companies to be 400 per cent over 3 years. Here's how the calculation was done

Cognite, an oil and gas / energy software company based in Oslo, produces the 'Cognite Data Fusion' product, which brings together all kinds of data from different areas of operations, to make it easier to make decisions with.

It sounds like a good idea, but the software is expensive to buy and install. Companies might want a sense of the return they will get from using it. So Cognite asked Forrester Research to try to work it out.

Forrester calculated that the return on investment (ROI) is 400 per cent over 3 years. This is based on creating a 'composite company' which is a mix of 6 actual companies studied. Forrester was given access to real financial data for these companies, but was not able to publish it. It can verify that the composite company's figures are realistic, without needing to betray any secrets.

The composite company spent $5.44m on the software over 3 years, including Cognite's license fee, the costs of maintenance and management of the platform, and its implementation.

Much of the value from using the software was from optimising its use of heavy machinery, reducing plant shutdown time, and improving energy and maintenance. The value gained worked out at $108m over 3 years.

The software did not directly cause the improvement, it was a tool which staff could use to make better decisions which led to the improvement. To account for this, Forrester attributed only 20 per cent of the gains to the software, so $21.6m. The 400 per cent number is calculated from 21.6 / 5.44.

Cognite brings together data from different places and puts it together into dashboards and for feeding into analytics. It makes it easier for people in the company to find and work with data, which means that they are motivated to find new ways to use it.

The research study was done both to understand the value customers are getting, and to help Cognite better understand customer priorities.

The movement fits with a broader trend of manufacturers shifting their priority 'from grease to code,' said Paul Miller, VP and principal analyst with Forrester, and a specialist in smart manufacturing and IOT. They want to wrap 'digital services, capabilities and insights' around their manufacturing.

It also fits with the trend of companies building 'digital twins' - digital versions of something in the real world, since integrated data about a real thing and a 'digital twin' are much the same thing.

They can use this digital twin to try to answer challenging questions such as, where is the best place in the world to manufacture, and how to best manage energy and achieve decarbonisation, Mr Miller said.

It connects with companies' efforts to bridge the gap between operations technology and information technology, he said. 'The IOT platform on its own is not enough, ERP on its own is not enough. To deliver real value, you need to start stitching these systems together.'

Companies cannot yet buy a 'digital industry platform' with everything integrated, but software is moving in this direction, he said. 'We're starting to see the pieces that will allow people to stitch these things together.'

The study

The study aimed to assess what Forrester calls the 'total economic impact' of the software, from capturing and quantifying all of the benefits. Forrester followed the same methodology it uses for hundreds of similar product analyses every year.

'It's a proven, consistent and repeatable methodology,' said Jan Sythoff, Principal Consultant with Forrester, and lead consultant for the research project.

Forrester staff began the study by seeking to understand the product as well as they could, and then interviewing six customers to collect their data.

With the customer data, it could build up a fictitious organisation with data which was something of an average of the six customers it spoke to, so the data is representative of a company in the real world, but without containing any confidential data. Then Forrester built a 3 year financial model for that 'composite' company.

The six real individuals and companies included a product manager from a European oil and gas operator with $250bn revenue; a VP data governance from a European oil and gas company with $6bn revenue; a data analytics manager from a US oil and gas company with $7bn revenue; a senior VP of a European marine geophysics company with $600m revenue; and a $3bn revenue utility and a $275m manufacturing company, both European.

The 'composite' fictitious organisation, developed by Forrester for the study, had $2bn revenue from 30 different assets, mostly manufacturing sites, equating to approx. $200k revenue per asset per day.

It had implemented Cognite Data Fusion in 20 assets in the first year, 25 in the second year and 30 in the third year.

Financial results

This 'composite' company did not see any benefit from Cognite's software in the first year, and its expenditure on it was high. But in the second year, the company saw benefits from the software providing staff with better access to data in general, then helping them find ways to optimise heavy machinery use, and reduce shutdown time. In the third year, it helped staff find ways to improve energy and maintenance. 'This was typical with customers we spoke to,' he said.

Each year from the third year onwards, the composite company gained $9m from optimising heavy machinery, $5.1m a year from improving energy efficiency, $4.8m from having shorter shutdown time, $4.3m from having an optimised maintenance program, $2.3m from real time data efficiency, and $1.5m from productivity savings.

The study of reduced shutdown time found that the fictitious company had 5 extra days of operation in the second year, and six in the third year, leading to additional revenues at $200k / day.

The study of maintenance optimisation found that 17 maintenance days were avoided for each asset from using the software.

The composite company saw 15 per cent energy efficiency savings. There were other savings which were harder to quantify, including from making experts more productive, making ESG reporting easier, and improving health and safety.

The software did not directly achieve the benefits, it was a tool which staff could use to identify ways to achieve the benefits. So Forrester's study attributed 20 per cent of the gains to the software itself.

On the costs side, the composite company paid a subscription fee of $3.2m to Cognite for its software over 3 years. The subscription fee depends on the number of assets the software is used on, the types of data to be 'fused' and the number of services deployed.

The 'operating costs' were $2m over 3 years, including for maintenance and management of the platform, some fees to external consultants for implementation and support. The one-off 'implementation costs' of $241k included costs of client company staff. The study was based on an initial implementation period of 6 months.

The net present value of the software based on 3 years use was calculated at $21.6m. So based on initial costs described above summing $5.44m, this means 4x return on the investment (or 400 per cent).

In the first year, the costs are higher than the benefits; in the second year, benefits start to 'kick in', and then increase in the third year.

There was no modelling done for subsequent years, but unless there were substantial software upgrading or customisation costs, there would be bigger returns and bigger ROI.

Problems and benefits

Interviewees commonly said that before using Cognite, they had multiple data repositories which didn't talk to each other, with data formatted in different ways, leading to incomplete dashboards and data analysis. They couldn't easily connect data from one asset to another, to compare them. Individual pieces of data meant different things on different assets.

Companies would often do 'point' use cases, putting a lot of effort to try to get one dashboard or set of data, but they could not re-use any of the work, so it was not very efficient. They were looking for a way to bring all of their data together into a single repository, and make the data easily available.

Cognite can help with much of this. 'We can see a 3D model of the facility and tie that to a database that monitors the condition of each piece of equipment,' one oil and gas company data analytics manager said in the survey.

One drilling manager said that Cognite was helping them improve performance of drilling, such as helping them make a better determination of the best weight on the drill bit, the rotation speed, and other drilling parameters.

An oil and gas data analytics manager said that use of the software has led to a reduction in downtime of 20 per cent, by helping identify opportunities to do work simultaneously during the shutdowns.

Composite digital twins

Integrated data about a large asset, as Cognite provides, can be considered a 'composite' digital twin. This is more than most digital twins today, which just cover a single piece of equipment made by one vendor, such as the manufacturer of a pump or motor, and do not provide any view of the machine in relation to other assets around it, Mr Miller said.

These single manufacturer digital twins do not have standard metrics, assumptions and APIs, which makes it very hard to put them together, Mr Miller said.

A composite digital twin can 'stitch these things together, build up a picture of an entire workflow / process, and all of the assets which come together to get that work done,' he said.

For example, for a new wind farm, you could imagine a composite digital twin of the entire farm, which could be used to optimise maintenance, production, operations, and commercial aspects of running the farm, he said.

Emissions data reporting

Software tools to integrate data can also be used for emissions data reporting.

Gathering scope 1 and 2 data are relatively easy, because the data is under your control, or easy to obtain, he said.

But Scope 3 data, particularly for your supply chain, 'in a surprisingly large number of cases, is frankly a guess at the moment,' Mr Miller said. 'We don't have the metrics, data, sensors, to build up an accurate picture.'

'Quite often, when you dig into someone's emissions, you find the guess work quite quickly.'

'That's not sustainable. We've got legislation coming through in European and US markets that require this to not be a guess.'

Advanced tech

Mr Miller was asked what impact advanced technology such as Chat GPT and the metaverse might have on industrial digitalisation.

For Chat GPT, 'large language models are very interesting,' he said. 'It does a first draft [of some writing] and someone with knowledge cleans up the quirks.'

'A number of my colleagues are looking at this more closely. They say, 'these tools are great if someone with knowledge is looking at what they are doing.''

'It wouldn't be wise to let any of these models run off on their own without adult supervision any time soon.'

We should not expect Chat GPT to solve big industrial problems of today, he said. If you develop an understanding industry 'pain points', particularly in safety, you can better understand whether the 'latest buzz technology' can assist with them. 'You often find it doesn't apply.'

And setting up a 'skunkworks' team to play with a technology is very different to deploying it on the factory floor,' he said.

Similarly, there is not yet any 'metaverse' for industry, Mr Miller said. Some companies are doing more with IOT sensor data, to help people get a deeper view about what an asset is doing, and developing more advanced digital twins, which could be a step towards an industrial 'metaverse'

'We're writing a report at the moment about the overlap between industrial twins and the metaverse. Its an interesting research area but don't get too excited by the hype,' he said.

You can watch the webinar online at
https://www.cognite.com/en/webinar/forrester-deep-dive-total-economic-impact-of-cognite-data-fusion

Associated Companies
» Cognite

CREATE A FREE MEMBERSHIP

To attend our free events, receive our newsletter, and receive the free colour Digital Energy Journal.