Data factories are necessary but not sufficient to compete with the leading companies. The disruptive tech companies did not become innovation leaders by building factories and data flows and then collected the gold. Data factories grew over time while these companies applied data engineering solutions to the product challenges in front of them. Driven by their immediate needs, they developed enabling technology, but also the ways of working that are necessary for data and AI success: cross-functional teams, data-guided product development, DataOps, etc. Changing ways of working is challenging for traditional companies, and failure to adopt new work patterns is one explanation to the high failure rates for enterprise data and AI projects.
Fortunately, there is a simple path to learn the necessary work patterns. It yields quick and high return of investment, and the technical leaders have taken this path, but most companies nevertheless choose other, more expensive and less rewarding paths. We offer to take you down the path to success that the leading companies have taken. We call it CorpOps.
At the technically leading companies, day-to-day operations are heavily automated, both technical operations and business operations. The core business processes, but also accounting processes, customer support, sales processes, etc. all have undergone many iterations of automation. Much of the toil and repetition has been removed from humans, freeing up their valuable time to things humans do better than machines.
In these “born-digital” companies, there are developers dedicated to directly supporting business operations, including processes outside company core business. One can find teams with names such as CorpEng, analytics engineering, or financial engineering.
In most companies, this a utopia out of reach due to lack of developers in the non-technical departments. The born-digital have an engineering-first approach to problem-solving that other companies cannot easily copy, for cultural and cost reasons. We call this “CorpOps” - continuous, relentless automation and efficiency improvement of daily corporate activities, through software engineering and data processing, driven and directed by business and product experts.
What if we told you that you can have a developer team sitting next to you dedicated to make your daily work more efficient? What would your business teams do if they had daily access to software engineers?
What is the connection to data factories, you might ask? The answer is that CorpOps seems to be the only successful path to reach to industrial levels of data efficiency. While most traditional companies jump on the data and AI train with large, dedicated efforts with new technology and strategy that will enable innocation, none of the technical leaders went down that path. Instead, they applied engineering to the business challenges at hand, which over time created both data lakes filled with relevant data as well as capabilities to build product features based on that data. From there, cross-functional teams with a mix of data engineering, product, and business domain experts organically innovated and thereby took lead in the data and AI evolution. Their great technical platforms and components were not strategic projects decided by top management, but instead grew incrementally from solving concrete business needs.
Scling’s mission is to enable companies to reach the same level of data innovation as technical leaders, so we offer to go down the same path together with you - automate your business proceses, collect data about your business, and grow your ability to improve your business with that data. We have built CorpOps business automation at technically leading companies as well as for our own internal purposes, and we operate a platform that enables us to quickly build custom automation.
You will need to contribute the puzzle piece that we lack - knowledge about your domain and your business. We will collaborate and improve your business together with your business domain experts. Close collaboration between software/data engineers and product or domain experts is a recipe for success in the disruptive tech companies. It is rarely seen in traditional companies, but we have worked in very successful cross-functional teams and will guide the collaboration.
For our own benefit, we have automated our processes to eliminate recurring tasks, both administrative and technical.
These task automations have quickly yielded return on investment, even though we are a small company. They have also been implemented without requiring extra information collection - they use the existing data that is already present in our systems: Gitlab, Jira, Gmail, etc.
The traditional ways of automating business processes fall into one of three categories, with different limitations:
Use an enterprise resource planning system or other general platform that aims to support many different businesses, customers, and processes. This gives you a wide reach across the processes you want to automate, more than you need. At the same time, you rarely get what you actually need — the implementations do not cover the needs that are particular to your business. We get a wide, but shallow solution. The general platform solution is also large, difficult to get started with, and does not deliver value until a critical mass of functionality is implemented and adopted.
Use domain-specific tools, for example automation for the accounting process. This has the upside that you now can automate most things within accounting, but leaves out the other parts of the business. The tools you pick for other niche processes have different models of the world, and you get silos of digital business information. We get multiple deep, but narrow solutions.
Hire contractors for a business automation project, implementing a subset of your processes from written requirements. At the end of the project, the system goes into maintenance mode. In business systems built in automation projects, the written requirements are comprehensive to cover most processes; as a result, some functionality is frequently used, but other parts are mostly unused, limiting the cost effectiveness. When processes change, low return on investment discourages iterative adaption projects. We get a customised, but static solution.
Whereas traditional business automation implementations are often viewed as a time-limited projects, technically leading companies instead approach business automation by having dedicated teams that know the business, are part of the business organisation, and approach the development of business process automation as evolving digital products. They continue to develop the automated processes, continuously improving them to remove manual work and to adapt as business needs change.
Off-the-shelf automation tools are difficult to adapt beyond their intended scenarios and therefore must aim to cover 100% of cases, which is both expensive and in practice not possible. With customised automation, we can aim for 95–99% coverage and separate out the remaining cases for human supervision. Such human-aided automation provides a faster return on investment, and we then proceed improving automation in iterations prioritised by benefit vs. cost.
The general platform and the domain-specific tool approaches do not fit into modern IT quality assurance processes. They are therefore unsuitable for use cases where functional quality or data quality is critical. In contrast, we apply modern software quality assurance with automated testing and operational quality monitoring, and can automate business critical processes without quality compromises.
We believe that efficient business automation that is both wide and deep should not be an exclusive right for the tech giants. We therefore offer CorpOps as a service - to be your development team for business processes. We have developed highly automated administrative and financial processes in born-digital companies, and have adopted the same practices for our own business processes. We have slimmed down the development and operations efforts to a degree that automation efforts yield return on investment even at small scale.
We implement CorpOps flows on our internal data platform, which is better suited to asynchronous automation of long-lived processes than microservice-based systems. It is optimised for frequent changes to business logic, it is naturally resilient to technical faults and human error, it preserves history, and it is operationally 10-100x more cost-efficient than typical online systems or off-the-shelf data platforms.
Preserved history from your automated business processes form a digital foundation for data innovation. From automation of existing processes, we can proceed to statistical process control of business processes, e.g., alerts for costs outside the normal range, and then on to risk predictions, forecasting or other digital business process innovation. This is how the technically leading companies started their journeys to digital domination — automation of mundane everyday business activities and enabling grassroots innovation in core business teams.