In this post I collect the best data – and UX-research – tools and I break them down into 3 categories:
1. The baby steps: this is for you, if you just start to use data at your startup/e-commerce company.
2. The standard toolset: when you have a little bit more money for your data projects and when you want to dig into more details.
3. The expert toolset: when you have dedicated data people and also want to have very specific researches.
Google Analytics: Of course the top well-known data tool. It’s for general data analyses. It’s very easy to setup and it gives you a very detailed overview about, what is happening on your website or in your mobile app. http://analytics.google.com – FREE
Google Tag Manager: It’s not for data analysis. More like put all your tracking codes into a clear infrastructure. Highly recommended to use it from the very beginning and implementing different services (like Google Analytics, Hotjar, etc.) with this. http://tagmanager.google.com – FREE
Hotjar: It’s an all-in-one user experience monitoring service. The main 3 features are: heatmapping (on any of your webpage), recording sessions (yes, you can see an actual video about your real users sessions) and exit-intent surveys (“why did you leave our website?”) http://hotjar.com – $29/month
Regex101: A small, but useful addition, if you want to user Regular Expressions with GA or Hotjar or anything else. It double-checks your RegEx code for you. https://regex101.com/ – FREE
The original 5-second-test: Five second testing is a powerful method, when you want to validate your design before coding. You show your design to potential users for 5 seconds, then asks questions like “What do you think this company sells?”. If you get the answers, you expected, you are doing it fine. If not, it’s time to redesign. This tool helps you to hire users as well. http://fivesecondtest.com/ – $1/users
Usertesting.com: So far the best remote usability testing service, I tried. You can hire users from anywhere in the world, and you can target very specific segments as well. You need to setup a scenario and tasks for the users, in exchange you will get ~15 mins recorded sessions, where the users are going through your website and they share all their thoughts. The best quality/price ratio on the field. https://www.usertesting.com – $49/users (for the first 10 tests)
Prezi: Wait. What? Prezi is a presentation software, not an analytics tool. Yeah, that’s true, but at the same time it’s really good for mind-mapping. I use it as my digital whiteboard. I love to put together all my important charts, heatmaps and usability test videos into one place – preferably with links and side-notes. Prezi is just perfect for that as it’s zooming and as it’s really easy-to-use and flexible. www.prezi.com – FREE(mium)
Excel: of course. When you have some raw data in .csv format, this is the right software to process it and turn it into charts. However I use Google Spreadsheets instead, because it has all the core features I need from Excel, but it’s in the cloud, it’s free, it’s connected to gmail, etc. https://docs.google.com/spreadsheets/ – FREE
Note: if you are really into Google Sheets, try App Scripts as well!
THE STANDARD DATA TOOLSET
Optimizely: The #1 AB-testing tool for your website (or mobile app). WYSIWYG, means you don’t even need to know coding to setup your first AB-test. You can split-test your landing pages, product pages, regforms, etc. I recommend Optimizely for marketers, too – to setup simpler tests (copies, colors, pictures). But make sure you get someone, who understand HTML, CSS, JS and other stuff, before you set up something more complex (eg. full layout changes, menu structure changes). http://optimizely.com – the price depends on your plan, but it will be a 3 or 4 figure monthly fee in $.
Crazyegg: A website heatmap tool, just like Hotjar. Crazyegg does not provide recordings or survey solutions, but it’s much more powerful on heatmaps, than Hotjar. First of all, you can use it on dynamic sites – eg. if you have a drop-down menu, you can do heatmap on that (it’s not possible with Hotjar). Besides you can break down your heatmaps by segments (by channels, by user types, etc.) and you can also connect it to Optimizely to generate heatmaps for AB-tested versions of your website. After a while these features are really needed. www.crazyegg.com – $99/month
Mixpanel: Mixpanel is like Google Analytics with one important additional thing: datapoints are connected to users. With that you are not only able to identify them, but you can personalize your campaigns as well. Would you like to send a push notification for all the users, who checked your product, but didn’t place it to the basket eventually? Do you wanna send an e-mail for all your customers, who bought more than 3 products and visited one of your specific landing? You can do that. www.mixpanel.com – 0-2000$+
Google Data Studio: A new tool from an established player on the market. I’m using Google Data Studio for 2 reasons. 1) It’s free. (At least for now.) 2) You can import any data source into it – so Google Data Studio became my favorite tool to compare different data sources. Eg. if I see different trends for the same SEM campaign in Adwords, in Google Analytics and in my own database, I can simply channel these 3 data sources into one Google Data Studio project and analyze the differences there (even on one line chart) instead of clicking around in my browser. This method makes it much easier to understand, which of my tools provides misleading information. https://www.google.com/analytics/data-studio/ – FREE
THE EXPERT DATA TOOLSET
As you can see, most of the stuff are free at “babysteps”. The standard toolset (Mixpanel, Optimizely, Crazyegg) can be more pricey and your costs will scale also with the number of your users. Beyond a point it’s worth to leave the easy-to-use graphic-interface analytics/testing softwares behind and start to build and use your own data-tools (scripts, pipelines, automations, etc.). Yes, it means coding! But don’t worry, it’s not difficult at all. Besides the costs (which will turn into developer-time anyway) these new tools come with 2 other cool consequences: your data will be more flexible – you can connect anything with anything. And you will own your own data – means it’s on your server. There are 4 common languages for data analysis:
SQL: The query language you should learn, even if you are not a data analyst. SQL is like Excel on steroids. It stores data in well-structured tables, but – compared to Excel – you can work on much bigger data-sets in a more flexible way. For that you should learn queries like this:
SELECT * FROM usertable WHERE usertype = “advanced”
And you should think a little bit more abstract as you won’t see the whole table with all the data all the time as it was in Excel. Also it’s a very strict format of storing data: if you want to use SQL, you should always plan everything really carefully before you act and think about everything before you setup your actual data warehouse. There are different SQL languages, lot of them are open-source and free. https://www.postgresql.org/ – FREE
If you want to learn SQL, start here: SQL for Data Analysis (for beginners).
Command Line: Command Line is not built for dealing with data, but happened to be a very useful language for that. If you are a startup, you change things fast. It means, your data-structure should change fast, too. So most probably you won’t log data in structured tables (like SQL), but in plain text format (csv, tsv or txt files), because it’s quicker and more flexible. In this case you will have a higher focus on clearing data before analysis. And command line is just the perfect tool to do that. But not just cleaning the data! As it’s an open-source tool, lot of contributors implemented easy-to-use analytics-commands into command line and it became a very powerful language for quick and dirty data analyses. (Note: Some of the bigger startup companies with hundreds of millions of users are using command line as well.) http://linuxcommand.org/ – FREE
If you want to learn the Command Line, start here: Data Coding 101 – Intro to Bash.
Python: Python is a very complex language. (Similarly to command line, it’s not just used for data analytics). It’s really good in processing structured and unstructured data-sets and in my opinion it’s the easiest language to learn for a beginner. If you set up Python, you will access the basic commands of it, but you can expand it anytime by importing new packages. And that’s the real power of Python: you can import machine learning packages, you can import advanced statistical-tool-packages. You can import everything, you can think about. Python is also very popular for text-mining and sentiment analysis. https://www.python.org/downloads/ – FREE
R: R is for math. Okay, this is not completely true, but comparing to the other languages above, this one is much more based on mathematical thinking. So if you like rules and consistency against the easier usage, you should go with R. You can import a lot of the statistical packages here, too, which makes R a really powerful tool to discover correlations, make predictions or just in general do data science. https://www.r-project.org/ – FREE
Tableau: If you go with data coding instead of the 3rd party tools, you will need a data vizualization software too. During my career, I’ve worked with GoodData, Looker, Chart.io, Qlik, Tableau and some other smaller vendors. For me the clear winner is Tableau as it provides the best value for money. It’s very user-friendly and quite popular – means it has a big online community and a lot of support materials online. But don’t take my word for it, try for yourself: they have a 30-day free trial. https://www.tableau.com/ – $42/month/user
So these are the tools, I like, use and suggest others to use as well.
If you are about to start with data, I suggest you to go with the “babysteps” tools. You don’t need to rush. You will know, when you grow out those and when you need to change to the standard or the expert set.
Remember: if you start immediately with the expert stuff, you will get confused by the millions of possibilities. And if you stuck at the baby steps, you will never learn to walk like a grown-up data guru! 🙂
Anyway: just make sure you know, what is your goal and that you pick the right tools to reach that goal.
I’ll dig into the specifics of these great tools in the next 2 chapters!
And if you want to be notified first about new content on data36 blog (like articles, videos, handbooks, etc.), sign up for the Newsletter!