So your startup is growing... Great! The only problem is, you can’t say exactly why it’s growing or how to increase the rate of growth. You know that analytics are important, but you don’t know quite where to start as far as building the operation is concerned.
Maybe it all sounds very complicated, or you’ve taken a look at the universe of analytics resources available these days and have no idea whether you need an open source ETL tool or an enterprise-scale, does-everything data management suite.
Don’t worry, we’ve got you.
At Panoply, we’ve helped startups from Saucey to Spacious build out their data operations, so we know a thing or two about the ins and outs of putting together an effective set of analytics tools for startups. We’ve put together a guide that covers how to get the major components of an analytics operation in place with a minimum of pain and expense, and it should be useful for most business types.
Startup analytics 101
First, let’s cover the major components of a data analytics operation. You’re going to need to assemble a set of tools that fall into three broad categories:
With these three core components in place, you should be able to build a simple, effective data analytics operation that can scale with your startup as it grows. Before we start to look at specific tools, it will probably be helpful to give a little more detail about how the components fit together and what each one will be doing in your data stack.
Data collection - what you need and how to get it
Unless your business is built specifically around highly specialized data and advanced data science techniques, there’s a relatively small set of data types and collection methods that you’ll need in the early stages. And even if you are planning on doing something extremely fancy, these recommendations will probably still apply for other segments of your business. The specific mix of data types you collect and analyze is going to come down to the type of business you’re running, but you’re almost guaranteed to need to be tracking least a few of the following:
- Ad performance: This will help you break down the cost-effectiveness each ad network, audience, and ad creative. It’s a must when it comes to scaling effectively.
- Web traffic data: Even the simplest websites generate a lot of insight about who is interested in your products or services. Besides, this will help you forever memorialize that traffic spikes you got from Product Hunt and TechCrunch.
- App tracking: You’ll want this type of data to understand how customers are using and interacting with your offerings, especially if you’re a software business.
- E-commerce data: This probably seems obvious—and it should!—but if tracking purchases and their associated data is your primary interest, you’ll want to combine this data with the other types to understand exactly how your customers go from finding your site to visiting your site to clicking the buy button—and where you an improve.
- CRM and marketing automation data: Data from your sales and marketing teams’ CRM suite will help enrich your understanding of your customer lifecycle, especially in combination with all the other data listed here.
- Customer support: You’ll also want to look at the flip side of things--how well are you handling technical issues and unhappy customers? This data often
- Net Promoter Score: Depending on who you ask, NPS is either the one KPI you need or an overhyped so-so metric, but since it’s such a simple data point to collect, you’re just better off tracking it than not. It can help you understand what’s making your customers happy (or unhappy!), where you stand with regard to other companies in that respect, and what might trigger your customers to leave.
- Financial data: You’ll track this internally, and again: integrating this data with the other types listed here will let you start to get better insight into how your business is performing, where it can be improved, and what to expect from the future.
- App database: Understanding what data is in the apps you use is also critical for the success of startup data analytics operations. Combined with the other data types listed here, this will help you start to build a full customer story.
Top data collection tools for startups
Now that you know where to start as far as the types of data you’ll need, let’s take a look at how to get that data. Here’s a sampling of the best resources for quickly getting started on tracking data about your business, as well as the types of data they can help you collect.
- Google Analytics: This should be at the top of your list—that’s why we put it at the top of ours. You can get started with Google Analytics by installing Google Tag Manager on your site. GA will help you understand traffic patterns and generate insights about how people are getting to your site, where they’re coming from, and even what they’re doing once they get there through the use of marketing pixels. If you’ve only got the bandwidth to set up and monitor one data source, Google Analytics should probably be it, because it’s just so useful. If you set it up right, you can use it as a tool to watch your customers’ entire process, from site visit to purchase. You can learn more on Google’s Analytics Academy and MeasureSchool.
- Ad data: Data and digital ad ops are two peas in a pod. Whether its Facebook Ads, Google Ads, Bing Ads or another platform, these platforms start generating data as soon as you click “Launch.” Keeping track of this data will help you monitor your marketing campaigns, and can be integrated with your Google Analytics data to get even finer detail on how your customers are finding you. Unfortunately, Google Analytics does not allow you to compare Facebook and Bing Ads data (that’s where a data warehouse comes in). And on top of that, things get really interesting when you tie cost data back to your CRM to calculate customer lifetime value.
- Snowplow / Heap: Tools like Snowplow and Heap will help you manage event tracking on your sites and apps, both on the server and client. They offer built-in analysis and reporting tools, but you can also pull in the data these services generate and analyze it on your own (more on how to do that in later sections). Heap is easy to set up and get running, so they’re good choices for a startup in the midst of building an analytics operation. Snowplow is a good choice if your engineers want more control over data collection and processing. It is easy to slot into a workflow that already involves Google Tag Manager. You can also check out Mixpanel if neither of these appeals. Hotjar is also a popular option for session recording.
- Stripe / Shopify: The data generated by Stripe, Shopify and other e-commerce platforms will help you better understand purchase behavior, and combining it with the other data types we mentioned above will help you develop a pretty granular understanding of the entire experience for your customers, from seeing an ad to making a purchase.
- Zoho CRM / Salesforce: CRM packages like those from Zoho and Salesforce are where you’ll go for the CRM data we talked about earlier. If you’ve already got a CRM suite you like, stick with it--your main concern should be the data you can generate, and keeping track of it. A good CRM tool for this type of data should allow you to look at your own sales reps’ performance and get a sense of conversion rates at the different stages of your sales process.
- Zendesk / Intercom: The data you get from tools like Zendesk and Intercom will help you track how your technology is performing, as well as how well your customer interactions are going. Track your tickets in Zendesk to see what types of problems are coming up, how well they’re resolved, and how often they’re coming up. Use your Intercom data to inform both tech support data analysis and customer recruitment analysis.
- Delighted: You can use Delighted to generate customer surveys that gather feedback and help generate NPS scores. Wootric is another popular option for NPS surveys. Whichever tool you use, this is a relatively easy datapoint to get from your customers, as it boils down to a one-question survey, and can help you figure out who your most satisfied customers are--and how to focus your efforts to create more satisfied customers.
Data infrastructure - managing all your data
While the data sources above are, for obvious reasons, crucial to setting up an effective data analytics operation, this next section really only becomes important once your startup’s data analytics needs grow past the initial stages.
In the very early stages, you can get by on a shoestring budget using some combination of Google Analytics, an app tracking tool, either Excel or Google Sheets, the built-in reporting tools from your data sources, and a lot of grit.
When you start getting overwhelmed by the sheer number of spreadsheets and built-in reporting packages you’re having to keep track of, though, it’s time to start thinking about setting up a data warehouse. Data warehouses have, until recently, been both expensive and difficult to set up, but things have changed! Multiple companies now offer inexpensive, cloud-based data warehouses that are easy to get up and running.
Depending on the data warehouse tool you choose, you may also need to get yourself set up with an ETL tool, which will allow you to pull in data from the various sources discussed above and load them into your data warehouse of choice.
Data warehouses for startups
At this stage, cheap, scalable, and easy-to-configure data warehouses are where you should focus your energy. We’ve put together a list of some of the best options in this range below:
- Amazon Redshift: Amazon Web Services offers Redshift, a cloud-based data warehouse that makes it easy to spin up a remote storage instance and integrate all your data in one place. It can be a bit more hands-on than some of the other options listed here, and is designed to work best with other Amazon products, so consider this if you have the technical resources and don’t mind pulling a lot of different tools together to make everything work. Want to give it a try? Here’s our tutorial.
- Panoply: Panoply adds a ton of convenient features to help with ETL, organization, and optimization. Panoply offers built in data connectors for every one of the data types and vendors we listed above, so it’s easy to load your data and move right to analysis and insight once you get set up. In most cases, you won’t need a separate ETL tool at all: just use our interface to connect Google Analytics, Salesforce, Mixpanel, Stripe or whatever other data sources you’re working with. Built-in integrations with a range of data analytics and business intelligence tools also make it easy to pull insights out of the data you have stored in Panoply.
- Snowflake: Snowflake makes another great cloud-based data warehouse. It can be a bit pricier and require more engineering resources than other options out there, but it’s a popular option for a reason. Snowflake has some data connectors, but you will need to couple it with an ETL tool before you start working with your data.
ETL tools for startups
If you’ve chosen to go with a data warehouse that doesn’t offer pre-built data connectors (or you have specialized data needs), you’ll most likely need a separate ETL tool to grab your data and load it into your warehouse so that it can all be integrated and readied for analysis. Some good, easy choices are:
- Singer: Singer is an open source platform that uses modular blocks to gather data from a source and load it into a destination database or warehouse. It’s great for building ETL processes on the cheap, and even though it’s a little more code-intensive than some of the other options here, it won’t be too difficult to set up even with a minimal level of technical skill. Check out our tutorial.
- Stitch: Stitch is an open source ETL tool that also has several paid tiers for users with higher-intensity needs. A solid choice for growing startups that know they need ETL tooling for their data analytics operations, but would rather have someone else handle the nitty-gritty.
- Blendo: Blendo offers a cloud-based ETL tool focused on letting users get their data into warehouses as quickly as possible using their suite of proprietary data connectors. Blendo’s ETL-as-a-service product makes it easy to pull data in from all sorts of data sources including S3 buckets, CSVs, and a large array of third-party data sources like Google Analytics, Mailchimp, Salesforce and many others.
- Fivetran: Fivetran is a no-code ETL pipeline builder with a broad base of pre-built data connectors, both for pulling data in and pushing it to its final destination. Fivetran works equally well for small data operations and enterprise-scale processes, so it can scale easily for a growing startup.
Analytics tools for startups
After you’ve got all your data collected, even if you’re not at the point where you find yourself needing a data warehouse, you’ll want to be able to do some analysis. There are a number of BI and data analytics suites that can slot into your data analytics operation whether you have a data warehouse in the loop or not, so you don’t really need to worry about future compatibility issues all that much at this stage. Some of our favorite data analytics tools are listed below:
- Mode: Mode is a great, highly flexible data analytics and BI suite that should be especially appealing for users who know their way around analytics already and like to tinker in SQL, Python and/or R. Strong reporting and visualization features and low price point make it a good choice for a growing startup.
- Google Data Studio: Google’s Data Studio is a free, open analytics suite that makes it especially easy to work with your data if it’s already in the Google ecosystem. That doesn’t mean you can only pull in data from Google properties, per se, but you will need to feed it into Google Sheets if it’s not coming from a Google-run data source like Google Analytics, Google Ads or Youtube. Data Studio is a great way to get started with analytics for small startups.
- Metabase: Metabase is a free, open source analytics suite geared at making straightforward, descriptive analytics tasks easy. That’s good, because the vast majority of day-to-day analytics tasks are fairly straightforward but can take a lot of effort to do and re-do.
- Looker: Looker is another paid service that has become pretty popular in the world of startup analytics tools. A strong set of data connectors, reporting functionality and visualization tools makes it a good choice for data pros who want a lot of control over their data.
So what’s next?
Now that you know what you’ll need, it’s time to get started! If the idea of setting up a ton of different data collectors, ETL processes and data storage solutions still sounds intimidating, you might like Panoply. Our wide range of prebuilt data connectors make it easy to load data from your data sources, eliminating the need to spend time setting up a ton of infrastructure. And with connections to all of the top data analytics tools, you’ll be able to start building dashboards, reports and visualizations in no time. If that sounds interesting, schedule a personalized demo today.