Survey identifies model-driven culture as vital for success in data science
While businesses are recognising the value of data science and its ability to enhance business revenue, implementing and scaling data solutions across a business continue to be a challenge. A recent survey of data and analytics professionals suggests that developing a positive business culture with employees is a major factor that influences the success of data science. Led by DataIQ, a memberships-driven forum for the data and analytics community, the survey covered a panel of leading professionals across multiple sectors and companies in the UK.
The survey findings showed that 1 in 4 businesses believe data science will impact their top line revenue by over 10%. The results also indicated a continued challenge with company culture, suggesting a positive, model-focused culture is difficult to develop and still needs to be focused 0n. Approximately 40% of respondents want more clarity of the needs from stakeholders and a further 38% understand the necessity to train business users in data science solutions. Furthermore, another 32% believe there is a need for a more positive relationship with their stakeholders.
Nick Elprin, the CEO of Domino Data Lab believes that most businesses begin their work in data science by employing several data scientists, but ignore the importance of developing a model-driven culture that corresponds with their needs and the needs of business users. Mr Elprin believes the survey highlights the impact of not having a positive culture has on identifying proper use cases, creating expectations and generating quantifiable impacts on the business. Recognising these challenges is vital for businesses so they can create the right path and scale data science solutions successfully.
Additionally, 40% of respondents indicated that limited understanding or support for data science in business is regarded as a major challenge. The survey suggested that 1 out of 3 businesses stated that the conflicts between IT and data science remain another challenge. Even businesses that regard their adoption level of data science and analytics as advanced are not necessarily free of cultural conflict. Other findings from the survey included:
Over half of all organisations believe they will experience an uplift of under 5%, indicating that the failure to implement data science contributes to lower expectations.
1 out of 5 companies is experiencing a significant competitive advantage via applying data and analytic tools in their organisation.
A total of 67% have assigned their data scientists together to form a core function, rather than dispersing them throughout the business.
1 out of 3 organisations believes they require months to get their models into production. This needs to be considered because market changes are constantly changing and models that utilise outdated data will not generate valuable recommendations.
1 in 10 businesses has implemented an enhanced automated monitoring model that creates proactive alerts when models are deteriorating. Data Scientists can then examine potential issues before they have any major impact on the business.
David Reed, the Knowledge and Strategy Director at DataIQ believes that for data science to provide real value, a positive culture needs to be developed, enabling stakeholders and data science professionals to collaborate and share common goals. Mr Reed explains that the survey results suggest that this is easier said than done. 4 in 10 businesses identifying limited understanding or support for data science in their organisation as their business challenge. This presents a cycle that results in 1 in 8 businesses failing to generate a compelling use case for data science.