Select a page

Organizational design for musedata

Organizational design is an important step for cultural institutions in becoming data driven and a key requirement to service any musedata themed strategy. According to a Bain & Company journal report on Big Data and Organizational Design, organizations with the best analytics capabilities easily outperform those without, a trend seen as early as 25 years ago in examples such as Proctor & Gamble. This same performance can be expected by the cultural sector with investment in data resource.

The introduction of analytics into the museum’s structure is a far more recent development, with the addition of analysis, science and insight leaders. This is largely dependent on the institution’s size, with a variety of design approaches explored. For all institutions, the biggest challenge of introducing a data driven culture is one of communication and education, touching all business units.

A pulse on organizational design for data in the cultural sector

According to a MCN conference survey reported on by the Arts & Metrics blog, 77% of responding museums already have a member of staff dealing with data. This year, the industry is seeing the beginning of dedicated data capabilities arising from next generation data driven digital strategies. For early adopters, such as the American Natural History Museum or the Minneapolis Institute of Art, this initial step is focused on appointing a data analyst role. In some institutions, such as the British Museum, this organizational design represents an initial foray into building a wider insights practice.

In most cultural institutions, the level of organizational development will be a matter of right sizing applicability. Once a museum crosses a threshold of around half a million visitors annually, the business case for a dedicated data analyst grows. At over a million visitors, the museum may begin to see more complexity in its data, requiring a more technically experienced data science role. For multiple venue institutions, those with several million visitors each year and institutions with a heavy data centric strategy, multiple data orientated roles will require the development of a fuller analytics capability as a team. In all examples, staffing can equally be a mix of inhouse capability, outsourcing, or both. Especially, this approach can assist access to deeper capabilities across a wider skill set using an on demand basis, a difficult mix to balance in a single role and a more affordable alternative than requiring justification of a dedicated full timer.

Importantly, museums need to discern between roles articulating data analysis versus data science. A data analyst will be domain specific, dealing with a limited number of data sets. For many museums, this first role will be limited to a web, social or digital scope. The analyst’s role will likely be reporting focused, working to slice and dice with tooling including database access such as with SQL, system reporting direct from source such as Google Analytics or analytics solutions like Dexibit. Data scientists bring a more academic approach with a higher degree of data transformation, advanced mathematics, machine learning and visualization development. Rather than only responding to insight enquiry or operational needs, this role will be strategic and architectural in nature, creating capabilities for others using development integration, algorithm manipulation and presentation layers. Advanced tooling requires development skill sets and might include languages such as R or Python, big data frameworks such as Hadoop or Spark and visualization libraries such as D3.

Either way, the role’s objective should be to do more than answer inbound data queries. Hire evaluation should concentrate on a candidate’s ability to problem solve, tell a story with data and contribute to strategic discussions. An ability to communicate effectively with data is paramount, yet difficult to master, as it requires an opposing mix of technical versus soft skills. This is a delicate balance between an analytical mindset, for example a discipline to lean sigma or split testing, versus organizational change management capability, with a knack for contextualizing the business value of data. Finding this technical and business balance in one individual will be a tough challenge, especially if the expectation is to traverse the executive conversation all the way down to detailed analysis. In larger institutions, the museum may need to appoint an executive level insights lead supported by analysts, scientists, or both.

In appointing these roles, a museum needs to uniquely understand the scope potential for the role. In the first instance, most will focus on the commercial imperative of audience insight using visitation data. However, opportunities for data science exist also in analysis of the museum’s collection itself and in analysis of additional data in support of curatorial activities, such as data journalism content for an upcoming exhibition. In a job market experiencing severe shortages for data roles, this extended opportunity for analysis is an attractive challenge for aspiring applicants.

Shaping the insight capability in the museum

A key decision in design is whether to centralize or decentralize data capability. A centralized approach will see a dedicated function established either reporting directly via an insight executive, or incorporated into the existing management structure such as via digital, technology or marketing departments. Alternatively, the capability may be spread at a unit level, with departments such as these, plus research, collections and curatorial, supporting their own roles. A hybrid approach suggested by Analytics Magazine can involve business unit data ambassadors collaborating in a cross department center of excellence. These sorts of collaborative working groups can also be especially applicable for managing data governance aspects across the museum, such as for privacy.

Regardless of approach, the individual or group’s mandate should firstly be to champion the data cause and act as ambassadors for the data driven museum. It will need to respond to service requests from throughout the museum organization, whilst establishing a central capability. The International Institute for Analytics recommends establishing a scope of service charter to set out these functions and advertise the group’s purpose to the wider staff. The group may find it helpful to use a responsibility matrix to highlight to others the decisions in the organization they will be involved in as consultants or data they will be keep others informed of.

An important factor to consider before embarking on new appointments is for museum management to consider how traditional audience evaluation research roles will evolve in a world with big data and analytics. Emerging industry trends may present researchers with career progression opportunities and newly developed insight groups will require understanding of existing evaluation methodologies, with a decision on whether this function is to be incorporated together with an insight team or managed in its traditional location with a clear partnership between the two.

Data leadership and organizational change

Whether the institution is appointing dedicated staff or not, a data driven museum will introduce the musedata theme for all decision making roles. As the Kellog’s Leader’s Guide to Data Analytics recommends, it is not enough that modern leadership will rely upon an analytics team to derive insight on their behalf; they must instead have a level of data confidence themselves. This will require leaders to know what to do with data, rather than just collecting it – encouraging managers to spend more time on deriving insight as opposed to spending time only on reporting. They will need to understand where their data comes from, what happens to it and how they combine this with their own domain knowledge.

Organizational design for musedata therefore doesn’t stop at introducing new roles or teams, but in adjusting existing departmental functions and job descriptions for a level of data responsibility too. New accountabilities will include involvement in data governance, capture, manipulation, reporting and presentation.

Much of this implementation will be concerned with cultural change, inspiring a metrics aware, listening museum with an agile development manner. To lead top down will require the museum to change its power balance and create opportunity for data points in the decision process.

Communications, education and training will become a significant challenge for the museum in facilitating this journey for existing staff and supporting new data specialists. Whilst data leaders will be able to assist others in learning, the museum will need to look externally to source development and mentoring for these specialists.

As with all organizations expanding analytics capability, one of the most significant challenges the museum will face is a severe market skill shortage in data resourcing and for data scientists, a hefty price tag. Setting realistic internal expectations during planning, advertising unique data research opportunities and exploring on demand shared services are all useful tactics in addressing this problem. Museums would be well advised to work in with industry training organizations and universities to promote the adoption of data analytics as part of the emerging museum professional curriculum and work with industry training organizations to see similar professional development opportunities for existing staff.

With 2016 billed as the Year of Musedata, this year will see exponential growth in analytics job and education opportunities as the cultural institution expands, with a proliferation of approaches and the beginning of a new data history to measure the effectiveness of each.