Through data, you are trying to avoid the biases and problems that come with making gut decisions. However, it is important to get your analyst briefing right to avoid unintentionally relying on an analyst’s gut rather than clear insights.
Many people think that data means only numbers but here it also refers to qualitative data, think customer surveys, reviews, etc. as that plays a big role in transforming your analytics function from graphs to insights.
You're reading a guest post written by a KPIs Studio contributor
Hey, want to have your article featured on the KPIs Studio blog?
Your Quick Access Links
Develop good foundations
A data program needs a good foundation. Without a measurement framework, data capture process and someone experienced in working with data you will get weak insights and low trust from the organisation.
One key part of your foundation is the skills & experience of your analyst. Data may be made of simple ones & zeros but actionable, relevant, timely insights depend on context & experience. This means analysts will often specialise in one area or industry. This may seem obvious but the assumption an analyst can interpret any data is pervasive & holds many companies back. Often you will find a large marketing team, management, creative & technical eCommerce teams, and one analyst trying to make sense of it all. As we will discuss later your brief needs to give them a running start, not make them start running in circles.
The second is the quality of your system and processes around how to collect, store, document, analyse, report, and disseminate insights. Data affects almost every part of the business and changes rapidly as tools and business practises change. If a company is not maintaining its data the foundation will break down.
Building on your foundation
With well documented, reliable data and an experienced person ready to analyse it you can start to turn data into insights and optimisation. Before we get into that, it is important to cover two things that stifle performance optimisation programmes.
The business must have a clear direction and be willing to adapt. Without a North Star to aim at the analyst will be forced to try and map the whole ocean. Without change there can be no magic in the data, any brand that thinks their market, customers, and organisation should stay the same has no need for analytics, only status reports. These two things often lead to dissatisfaction with the insights programme and high analyst turnover.
You will also want to think about how your broader decision-making process to avoid falling prey to Illusory Truths. With all this in place, you can start to think about what you want from your analyst and normally this starts with a question.
Types of questions
Analysts are not mind-readers; the purpose of a brief is to establish what the analyst is trying to help the company achieve. Often people ask questions that have no real yardstick to measure by so get a response the analyst thinks they can support if questioned, which is raw data.
Examples of these types of questions are: how is my site doing, is our marketing effective, what happened last week?
It is like a complete stranger saying, how did I do running across that field? The best you could do from a data perspective is to tell them how fast they went as a rough guess.
If an analyst receives these types of questions, without explanation, context, or an end goal they’ll probably spend a lot of time pulling a lot of data in the hope that some of it will stick.
The Why & the How
If it is very early days, you can create a Customer Journey map & benchmarks but then you need to improve and that means you are looking for the Why & How.
The Why is the insight into the finding. The How is the Recommendation.
Recommendations come in many forms but the main two are how to improve performance + what additional data is needed to form a recommendation.
Remember, recommendations for how to improve performance are rarely as straightforward as the requestor hopes for. Analysts are not generally in a creative or performance role. They can generally only measure what is there.
If one campaign is working well, the other badly, they can recommend shifting the budget to the higher-performing one. If there is only one underperforming campaign there is often no one, easily changed reason.
The best way to reduce this is to write down all the variables about a Campaign or launch. Then, over time, learn what variables have what impacts. Before this, there is too little for an analyst to make recommendations with confidence.
Recommendations to get more data are often put to the bottom of the queue as action and getting the next campaign ready is prioritised. This limits the impact of insights. Websites and website pages have many moving parts to be tracked. Ad campaigns have multiple channels and creatives.
Always build and expand your foundation, the best brief + analyst in the world can’t magic insights from empty databases!
How to craft the analyst briefing
Provide background and direction for the analysis.
We are focused on growing our profits without requiring additional sales.
Provide information about what is trying to be achieved.
The more specific the better. Break large briefs down into smaller ones where necessary. Use Metrics & Dimensions where appropriate to help reduce guesswork.
Increase the Purchase Transaction Conversion Rates from Mobile Sessions Landing on our Homepage.
Frame the parameters and key areas of the analysis.
What was the impact of adding a pop-up on the 6th Feb. The last major update was on the 15th of Jan (See linked Calendar for details of changes).
To understand what is having the biggest effects on performance from a Traffic source and New/ Return customer perspective:
- Are there any obvious issues with the Conversion Rate?
- If so, what?
- How confident are you and why?
Insight & Hypothesis
Provide business insights and context to focus the analysis.
- The Products we show on the page don’t entice our users because they are not relevant to the Ad that brought them there.
- The Checkout Flow breaks on some Mobiles/ Browsers.
- The page loads slower than average causing users to Bounce.
- We recently shifted to a more aggressive Acquisition strategy and are maybe getting lower quality traffic.
- We are thinking about changing the menu as we think users struggle to understand the groupings.
- We cannot change the information required in the checkout or product pricing.
- We have the capacity to test pop-up designs and Product pages. The category page design is fixed for the next two months, due to backlog bug fixes.
What response are you looking to get from the analyst?
This will help them to understand how long the analysis should take and how in-depth they should go. Are you looking for some quick stats to share with your team or in-depth insights and recommendations to be presented to a diverse audience?
Please provide your recommendations in slide format with a one slide summary for sharing.