top of page
Search

EdTech Marketing Guide #5: Marketing automation - Ellis Crosby @ Blinkist

Updated: Dec 12, 2019


EdTech marketing guides

Practical tips from leading EdTech CMOs

Purpose of these guides

Marketing is such a critical function for an EdTech start-up and we know that getting reliable marketing insight is not an easy task. In addition to The State of EdTech Marketing Report that we published last year, we decided to launch a series of short practical guides that we created alongside leading EdTech marketers.

These guides are intended to not only recognise outstanding marketers talent, but also to help you craft your marketing strategy, stay informed and to inspire you. Think of these guides as an EdTech marketing resource that you can tap into whenever you are in need of high-level tips.

We hope you find them useful and don’t hesitate to reach out with any questions and/or suggestions (We highly appreciate feedback at Brighteye!).


Join our EdTech marketing community

If you're interested in joining our curated EdTech marketing Slack community, in which EdTech marketers are able to exchange, ask questions and share experiences freely whenever they need it (e.g. advice, tips, recruitment, news, tools, etc.). , email us đŸ“© in order to receive an invitation.


 


The 4-step framework that will take your marketing automation strategy to the next level


- Ellis Crosby, Senior Marketing Intelligence Analyst at Blinkist

⬇ Download Guide ⬇





Blinkist is a leading book summary platform that provides key takeaways from bestsellers, new releases and recommendations. They carefully curate the books they choose based on reviews and through readers’ input, and most importantly, the summaries are made and polished by humans. They currently have 11 million users, and their mission is “Big ideas in small packages. Start learning now.”


Who’s Ellis and what’s his background?



Ellis is a Senior Marketing Intelligence Analyst at Blinkist and he’s in charge of handling all marketing automation requests to make everybody’s life easier and to improve the efficiency of their marketing strategy. Prior to joining Blinkist in October 2018, he spent some time as a performance marketer on both sides of the table: first in the agency world helping clients, and then in a more operational way, when he joined a company. He has always had an affinity for numbers and data, so when he’s not working, he spends his time on side projects around marketing automation & data. He recently started helping a digital school in Kenya as a charity hobby! 🙌



What motivates you as an EdTech marketer?


It’s pretty straight forward: contributing to building a better society. The company I worked for before Blinkist is a nutrition company (i.e. promoting protein shakes) and I did not really feel I had a positive impact on the world. Whereas, at Blinkist we are a lifelong learning solution, so we are allowing people - especially adults - to keep on learning. I feel that these people are then likely contribute to the rest of the society. For instance, when people acquire new knowledge on a specific topic (e.g. climate change, mental health, business, etc.), then they tend to share those new ideas with their friends and other people. To me, this is how you can make the world a better place and I really feel that there’s no harm in giving people information!



How is Blinkist managing its marketing automation efforts?


Since we all understand the power of automating marketing processes, in the Business Intelligence unit at Blinkist we receive tons of requests to automate specific tasks with the aim of freeing up time for our marketing team and allowing them to focus on valuable activities. We receive requests on a daily basis that vary from creating dashboards for tracking key metrics, automating a bidding process in a specific platform, tracking engagement in the app, to improving conversion throughout our funnel. So in order to be as efficient as possible in handling these requests, we developed and use a simple 4-step framework that helps us manage our entire automation strategy. I also strongly believe that this framework allows us to get to a stage where the automation becomes very powerful.


#1 Purpose & Coaching 🙋‍♀

Usually marketers come to us with a high-level solution in mind that they want us to build for them (e.g. tracking engagement in the app). This not always helpful, as the scope of their request tends to be broad. So the first step is to align everybody and to make sure we all speak the same language. We -BI- try to act as coaches: we sit and discuss their problem, understand what they are trying to solve and most importantly, what is the underlying goal.

Let’s take a quick example: you receive a request saying that “I’d like to track engagement in the app”. After asking the right questions, you discover that the true objective here is to track retention of users after X months using your app. So once you come to realise that engagement is not the key metric to track but rather retention, you can define with your stakeholders what the signals are for user retention, and run analyses to see what elements are correlated to retention. Even thought this first step seems basic, spending time with the stakeholders allows us to (1) understand whether the request has a strong purpose, (2) frame the problem, and (3) quickly assess whether we should move forward with it or not.


#2 Prioritization & impact 🔬

Once we have understood the nature of the request and believe it is relevant for our marketing strategy, we add it to our backlog and we define criteria that will help us prioritise all the requests. At Blinkist, It is extremely important for us to understand where and how to spend our time as we are only 3 analysts shared between 130+ stakeholders. We must focus on automating tasks that will unlock the most value, so we need to estimate (as best as we can!) the bottom line impact of this request for automation by understanding the potential impact: Will this project save time? Will it increase overall productivity? Will it help to generate more leads? Etc. Ultimately, our north star metric in the BI unit is impact in the business, so we must keep that in mind when assessing different problems to work on. Also, it is by far more impactful to spend more time refining a relevant automation request, rather than automating just for the sake of it. Carefully prioritizing automation requests is essential.


#3 Analysis & preparation đŸ§Ș

We deep dive by spending time with stakeholders and defining milestones. First, we observe their current process and way of doing things (e.g. what tools they use, what are the main KPIs, how manual is the process) so that we can build on it and see where the inefficiencies are. This way we understand what is truly automatable (note: only automate things that you’ve worked on manually!) and what information we -BI- need to make it happen.

The interesting part is trying to get into your stakeholders’ heads, by asking specific questions (e.g. “What would make you increase/decrease the bid now?”, “Do you look at this set of data #1 or set of data #2?”). We must understand their behaviours and decision making process. Secondly, we need to make sure the data is in a perfect state so that we can start building the solution. Let’s take an example: assume we want to automate a bidding process on a specific channel in order to increase the likelihood of conversion. To do this, we must (1) define which data we need (e.g. demographics, user’s device, time of day, location, time spent reading the ad), (2) define how to get access to this past data, and (3) clean this data before being able to analyse it and draw conclusions (extract data signals to inform future bids) from it.

Note: It is pretty common to encounter issues during the cleaning part such as not being able to get the exact data to the level required. This is where creativity comes into play and you must come up with new ideas and ways of tackling this issue.


#4 Delivery & Tweaks 💁‍♂

We are now ready to start working on the ad-hoc solution on our own. Depending on the nature of the solution we are building, we like to do quick catch-ups with our stakeholders to make sure that we are going in the right direction. One hint is to make sure the data is very clear and the stakeholder understands the process in detail. Once we have built the solution and it is approved by the stakeholders, we either A/B test the new solution together or we simply implement it to see it in action. We expect them to provide us with continuous feedback when the solution is live – the objective is for us to prove our hypotheses to be right or wrong; When we miss some of the hypotheses (we sometimes do!), we reflect on it (“What did we miss?”, ”Are we using the right data points?”, etc.) and we keep tweaking the solution until it works and fully meet their needs. In this last step, the stakeholders’ input is paramount as they are the ones who best know the process, and ultimately will be able to tell whether it is efficient or not. Bear in mind that the end goal is to track the impact of automation on key metrics and maximise your marketing’s performance.



 

In summary, these are the essential tips from the Blinkist’s framework:

#1 🙋‍♀Purpose & Coaching

Act as a coach to frame the problem and assess its viability. Avoid FOMO and cool/trendy, only work on automation projects that make sense for your marketing strategy.

#2 🔬 Prioritization & Impact

Spend time defining the potential impact of the automation on the business’s bottom line. Automate only processes that you have done manually because you really need to understand everything that is going on and where the value is.


#3 đŸ§ȘAnalysis & preparation

Spend time with stakeholders, get into their heads to understand their behaviours/decision making processes. Be friendly and make sure that everyone speaks the same language. And make sure you have clean data.

#4 💁‍♂ Delivery & tweaks

Prove your hypotheses by receiving continuous feedback from stakeholders and keep adjusting the solution. Give yourself room to adjust and fail safely because you can’t really understand the full problem until you start doing the work. You will probably fail a lot so be prepared.

⬇ Download Guide ⬇





bottom of page