So you have got a job as an analyst in your dream company? Here are some helpful tips to ensure your career gets off to a good start.
Learn the business – Analytics is largely concerned with building models. Regression, trees, neural networks are all techniques to build models. A model is just an algorithm or a group of rules that connect a set of information (input variables) to a particular target. There are two essential ingredients that go into the making of a good model –
Knowing the business plays a big role in how well you leverage the above two ingredients. For instance, if you are assigned to the team that serves the FMCG industry, it is very important for you to understand how marketing, pricing, sales, promotions etc. work in this industry.If you are evaluating the effectiveness of a marketing campaign for an FMCG product, domain knowledge will help you identify the 10 or 100 variables you need to look at from the thousands available to you. Domain knowledge will also help you interpret the results of the model in the context of the specific business problem. Knowing the industry jargon is essential when presenting the results. The results of the analysis have to be presented in a language that client understands or he will not be comfortable with your recommendations.
In some cases, companies may have an ‘industry-101’ which provides introductory information. The internet is a great source of news and other kinds of information. There are numerous whitepapers available online on all big industries. Finally, there is no better way of gaining industry knowledge than through talking to people with the relevant experience.
How many industries do you need to know about? The answer in most cases is ‘one at a time’. While you would get opportunities to work in multiple industries, you are likely to focus on one when you start off in analytics. If you are working for an in-house unit, your domain is the industry of the parent business. Example, someone working in HSBC’s captive center is likely to work on a domain within financial services, like auto loans or insurance. If you are working for a boutique analytics company that offers analytic services to clients in multiple industries (ex – Genpact or Musigma) you are still likely to be assigned to client(s) in a particular industry. Make full use of this time to build the business knowledge that will be your differentiator further down in your career.
Know the client – Knowledge about the client is also very essential for much the same reasons mentioned earlier. Anything going on with the client could help explain what you are seeing in the data. While analyzing the sales in the last 3 months for a retailer, a steady downward trend was noticed. The team struggled to explain the reasons behind this trend. Till they ran into a news article about the racial discrimination suit filed against the company by a few of its employees. A quick investigation into the geographic spread of the affected regions and a comparison with the population break-up by race confirmed the hypothesis that the news had generated a negative sentiment against the company within a community thus leading to the falling sales.
Again, the internet is your primary source of information. Set google alerts for any news on the clients you are working on. Get as much information and news as you can about them. It could help you in unexpected ways.
Be good to your peers – Analytics is a vast subject and no one can work on everything.Knowledge is usually fragmented within an analytics team with certain people being experts in certain industries and or analytical techniques.
Learn the tools – There are a variety of off-the-shelf analytical tools, free and commercial, available in the market. Many of them are being used by various companies in India.
In addition, you may need to work on different in-house tools in use by different clients. Spend time to understand the features and capabilities of the tools you would be working with. While you may not know everything about the tool, you should have a very good idea of all the things it can do. You can dig deeper into individual features as and when required.
A good knowledge of the tools will make you more efficient at your work and less constrained in your choice of technique. Excel, SQL, SAS, R and SPSS are the most widely used tools in the industry. A working knowledge of these tools at the start of your career can be a strong differentiating factor for a new analyst.
Focus on presentation of results as much as the results itself – People tend to get too involved in the complexity of analytics. They feel it is important to throw statistical jargon into the presentation in order to impress the client. The client is never impressed by complexity. The client wants simplicity. He is looking for simple, intuitive explanations that broadly fit in with his business experience.
Try to simplify your findings. A model with fewer variables is not only easier to understand, it is usually more stable too. Be prepared to accept some loss in accuracy as cost of simplicity. A model which explains 95% of the variance and is understood by the client is far easier to sell than a model with 98% accuracy but very low interpretability.
Present your results in a language that the client understands. While you need to exercise restraint when using statistical terminology, feel free to use as much industry terminology as you want. This is the language the client understands.
Spend as much time on the presentation of results as you do in getting them. Use fewer variables in your final models. Use the industry terminology in interpreting the results. Make your results easy to understand for the client and half your battle is won.
A good foundation is essential for a successful career. Spend time on the right things at the start of your career and things will become easier as you progress further. Always remember, analytics has little or no relevance without the business context. Focus on the business and you will succeed within analytics.
Click here to read the Beginner’s guide to analytics.
Click here to learn analytics.
Click here to learn about analytics courses.