Outsourcing Marketing Analytics

Analytics is increasingly becoming more and more important for marketing departments. There are more channels, more data points and more tools to deal with. As the complexity increases, organizations are looking at working with an external agency to deal with the analytics function. This article looks at some of the common scenarios where outsourcing happens, the challenges with outsourcing and best practices to deal with these challenges.

1. Lack of time

We see this time and again with our customers. There is always more work to do than what’s possible. Marketing analytics has not reached the stage where a senior analyst can start doing analysis without going through the data collection-processing-presentation cycle. Letting an agency handle part of this can free up time for the marketing department. In this case, outsourcing is done not for lack of skill, but merely because of lack of time.

 2. Lack of expertise

 Marketing analytics requires a range of skills and there are times when a particular skill is missing. It is perhaps too expensive to fill that gap or at times the need for the expertise is ad hoc in nature. Some of the examples we have seen specifically in the digital analytics industry are:

  • Building a solution design document.
  • Creating a predictive model
  • Building a data warehouse

3. Greater flexibility in costing

Things are moving faster and faster within marketing analytics. Newer tools, data sources, methods, engagement models and channels make it harder for marketing departments to keep pace with these moving pieces. Some of them want to have the flexibility of keeping the people cost variable. Working with an agency is one way to accomplish this.

4. Getting past the ‘hiring freeze’

As companies go through difficult times, there is usually a hiring freeze across the board. A marketing department that wants to hire an analyst is also impacted by this hiring freeze. Working with an agency at times is a practical option for companies under this scenario.

5. Need to focus on core areas

Many a time, operational areas take away the time of the marketing department from the core, strategic areas. If a team choses to move focus from operations to strategy, working with a service provider is a good option. This allows marketing and analytics teams to spend more time on things that cannot be outsourced.

6. Need for speed, economy and scale

An external agency typically would provide the marketing analytics function at greater speed and lower cost at a larger scale. The agency is built to provide marketing analytics as a service and can operate at a higher efficiency, which in turn could result in lower cost and faster turn around time.

Sometimes, the need to scale for some companies is of paramount importance. Working with an agency usually allows companies to move faster and scale more efficiently. The corollary is also true. Companies at times need to scale down faster as well. Working with a service provider allows companies to keep this option open.

7. Need to get an external point of view

There are times when it helps a company to get an external, unbiased view from someone who is independent.

8. Need to understand best practices

An agency by definition works with multiple customers and a range of projects making them well versed with best practices and best practices. Companies frequently approach agencies to benchmark their operations and bring in best practices.

Challenges and pitfalls of outsourcing marketing analytics

Any kind of outsourcing comes with its own challenges. Communication breakdown, mismatch of expectations, role clarity, etc. These challenges are true for marketing analytics as well. But here, I will try to touch upon some of the not so obvious challenges and pitfalls.

1. Communicating the context to the agency

When an agency works outside the organization, getting the context gets difficult. Without the context, an analyst would not be making optimal conclusions. Context could mean any of the following:

  • Organization changes
  • Change in priorities
  • Specific marketing efforts initiated
  • A PR event that was significant enough to impact the numbers

2. Integration with other departments

Departments within an organization tend to be integrated at some level. It could be because they share the same physical space and get to interact with each other. It could be based on access to common resources like an intranet portal or company-wide communication. An analyst who is outside the organization would be less integrated with other departments and could miss out some of the insights that come from this integration.

3. Retaining the core knowledge and avoiding dependence

Analytics is becoming increasingly core to organizations of all kinds and sizes. When this function is outsourced, there is a danger that the core knowledge revolving around analytics is outside the organization. This in turn could also lead to being dependent on the agency.

4. Figuring out how to incentivize the agency

Incentivizing an employee is a fairly common practice and a method that many companies have figured out to a larger extend. Company goals are broken down to various departments and finally to people. However incentivizing the agency is different. Buying hours / resources is the simplest way to start outsourcing. But how do you incentivize the agency to look for efficiencies. How do you reward the agency for optimizing processes? Performance based mode is at the other extreme of buying hours. Here the challenge is ensuring that the agency has reasonable influence over the outcome.

5. Giving agency access to data and tools

This is probably stating the obvious, but is worth mentioning. If not planned properly, the entire engagement can be at risk. Various departments including the IT department would be involved in giving access to an external agency. Giving access to tools that are in the cloud are relatively easy. But giving access to data and tools that are within the corporate network requires permission and more importantly time.


Recommendations and best practices.

If you can articulate the problem, there is a very good chance of solving that problem. So based on the challenges and pitfalls we have seen, here are my recommendations to make outsourcing of marketing analytics work.

1. Agree on a communication protocol

I can’t stress how important this is. Agreeing on a communication protocol will bring in the discipline of communication frequently and adequately. Depending on the nature of project, you can arrive at daily / weekly / monthly / quarterly frequency for calls and meetings. Frequent communication can help you communicate the context better to the agency. It would also help the agency to be better integrated with your operations.

2. Give access to the agency to internal portals.

Depending on the size of your organization, this might or might not be possible for you. But if it is possible, intranet portal is a great place for an agency to understand your organization well, know what’s happening, what kind of departments exist, etc. We have found in our experience that access to some parts of the portal help us understand our customers’ business better.

3. Think security from day one

Be prepared to spend time and energy on this. Plan ahead and speak to other departments like IT and legal. If some other departments have worked with external agencies, they would have tackled the security problem. Learning what worked for them would make it easier for you. I would also recommend factoring in an appropriate time frame to deal with security so that both parties do not get frustrated waiting.

4. Define the documentation that would be provided

Irrespective of the complexity of the job being outsourced, agree with your agency on the kind of documentation that would be provided. This would help you to retain the core knowledge and avoid dependency.

5. Build incentives

Once your engagement with agency stabilizes, it would be good idea to explore building incentives for the agency. From our experience, we would not recommend doing this from day one of the engagement unless you are extremely clear on how the engagement is going to pan out.

6. Identify a stakeholder who will own the relationship

We have seen better quality of engagement when there is a single stakeholder at the customer’s end who owns the relationship. This person would have access to people and departments who can make the analytics team more effective. This stakeholder can bridge the gap between the agency and your organization well.

7. Work on giving a good brief

A good brief and vision document does wonders. This can be in a written format or even a call. But taking the trouble of explaining your business, your objectives, your challenges and how it ties in with the task in hand helps the analyst get a larger perspective. An analyst within your organization might pick up this information automatically. But when the analyst is sitting outside the organization, it helps to give a good brief.