Analytics is more crucial than ever in today’s digital marketing world- but it’s about more than cold, hard data. Sandeep Saini, VP of Analytics Grazitti Interactive, a digital agency, and Chiara Pensato, Senior Director, EMEA Marketing, Alteryx share why connecting with the rest of the business is a crucial skill for analysts.
We’d like to share some information about omni-channel marketing – and about leveraging data analytics tools to do this well.
But before we jump in, we wanted to share a story.
Sandeep was hiring for a business analyst role in marketing analytics at eBay. There was a lot of interest for this position and he received many résumé. Now, these were some of the smartest people who had attended top universities from all around world. Every resume was better than the previous one – yet there was something missing in most of them. Due to technical jargon, it felt as though he were purchasing a software solution rather than meeting the person behind the CV.
In our day-to-day life we have become a résumé that describes all the complex problems we have solved and the tools we have used to solve those problems. We tend not to think about how many lives we have touched, the contributions that we’ve made to make the world a better place, and benefits that we have brought for the environment…
For example, the opportunities created by eBay and Etsy for small town artisans, who hitherto sold only in local markets, and have since had the world opened for them. Or YouTube, which, lets any artist share his or her music album to fans all around the world with no delay or gatekeepers between artist and fans. And Twitter and WhatsApp let people connect in seconds and unite over global injustices or family matters alike, cost free. All these examples rely on a solid foundation of data and technology to make them work at a global scale.
Data insights are built to support business priorities
The quest to democratise data and allow all decisions be made based on data-driven insights is an important goal in marketing. But, often, we as analysts fail to do that. How often have we worked on an interesting problem that we believed would change the way we do business, only to have it fall on its face, and not change anything? How many times have we complained that our audience does not understand our point of view or that they are too old school and against data-driven decision making? The question must be, in this data-driven age, have we as analysts done the required soul-searching and attempted to understand what the root cause for this failure is?
Going back to Sandeep’s task at hand to hire an analyst for eBay, almost every single person he spoke to was very enthusiastic and took pride in sharing how they built complicated models by ingesting hundreds, if not thousands, of variables.
The one thing that he almost never heard was how they sat down with business units and tried to understand what it is that they were trying to solve. None of the analysts ever spoke about any discussions that they had had with marketing research and social media teams, or with focus groups, to understand the customers and how the brand or products were perceived by them.
The most sought after analysts on the team were the folks who took pains to connect with the business. They attended planning meetings with leadership, worked jointly towards identifying the KPIs for the success of the business, and went out of their way to communicate the analytics plan in order to get everyone on board. It’s a sure-fire way of building a successful project.
Given Alteryx’s independent research findings that across the UK and US the process of making a decision, from requesting business information to the moment of decision, takes over 12 days, there’s clearly something to be said for every aspect that give an analyst an edge. The whole UK decision-making process tended to run a little longer than the transnational average, taking just over 14 days. This is why it’s so important to have the project and stakeholder management side sorted, Chiara believes. Having great models built on clean data is one thing, but the best insight won’t land as effectively as it could without decision-makers that know why it’s important and how to use it. That’s where the soft skills come in.
Getting business leaders on board as supporters
- The first step is to connect with your business leaders on a personal level
- Next, try to understand:
- What are the big challenges on their plate that data can help solve?
- What are they responsible for driving in the current quarter, half year, and year?
- What are their annual targets?
- Take those goals and map out what you are working on to see if they can align
- Dumb down the analyst jargon – these leaders aren’t part of the same world, you have to join theirs’
- Break down complex ideas by explaining them with real life examples. Help them understand the importance of statistics
One of the reasons we as analysts fail is because we rarely take our time to set expectations with our stakeholders. Having a heart-to-heart conversation where you explain to them that this is a journey, and not a destination, is very important.
Far too often the business has an expectation that is completely unrealistic. Sandeep vividly remembers when a manager at a previous job walked up and said, “Sandeep, do not even think of running an A/B test unless it improves the conversion by 50 per cent”. He went home that day and told his wife that it was time to look for a new job. If it were possible to improve results overnight for a business that took years to establish, life would be great! In actuality, life is not that neat.
That’s why you need to bring everyone to the table and help them understand that in real life data is messy and incomplete. The analysis that was done on perfect data and resulted in identifying the persona of every single buyer only looks so good during a demo. Data will be far from clean in real life. As an analyst you will have to cleanse it, enrich it, remove the outliers, and make it ready for consumption.
Luckily, there are new platforms which help to reduce the significant amount of manual work that analysts had to do in the past to ‘munge’ the data into something that can be usefully worked with. Rather than relying on legacy systems of static data warehouses, analysts can use software to pull information from a wide variety of sources into one central workspace, whether it be Excel spreadsheets and XML files stored in a company’s data warehouse, huge data sets stored on Hadoop, or publicly available data sources from the Internet. Once all the necessary information is present on one platform, rather than spend the majority of their time preparing their data, analysts can easily pick from their resources and instantly blend the relevant information together to create a workflow specific to the business problem at hand. Solutions that allow a flexibility of sharing, and facilitate cross-department collaboration can help analysts break out of siloes and orient themselves to solve problems across the entire business.
In our opinion, a one-size-fits-all approach never succeeds. There is an analytics path for each one of us that should be built one analysis at a time. In discussions with business leaders, many ask ‘which attribution model is the right model’?
Our take on this question is this: When you go to a restaurant they have 20 different items, not everyone is supposed to eat the same dish – which is why they offer 20 options. You ask your server for a recommendation based on your liking and yet at the end you are the one who must make the choice. Likewise, there are recommendations for attributions, however, what is the ‘right’ model can be very hard to answer.
Let’s say your organisation is new to attribution. The team is not very savvy, and the cycle time between lead creation and closing is comparatively short, so you might be okay with a first-touch attribution. On the other hand, you have a very long lead life cycle and as a company you have made a conscious decision that lead creation, opportunity creation and closed/won are equally important milestones in this journey. Then perhaps a W attribution is a good fit. There is no one-size-fits-all in this case.
To be successful in the role of an analyst, always put yourself in the shoes of your business partners and internal customers, and see that all your goals are aligned with theirs and that they speak their language. It’s a good way to become indispensable and secure a lot of colleagues competing for your talent.