Your Shopify Brand Doesn’t Need a CDP

Customer Data Platforms (CDPs) have become a growing topic of conversation with the brands I work with, with many platforms selling themselves as the single solution to all the brand’s data needs.

The grand promise of a CDP is to bring all the data points your brand holds about a customer together into one ‘golden profile’, and then allow better segmentation, audience building, analysis and personalisation.

This sounds great in theory, but in practice many brands who implement a CDP never see the benefits. Many CDP implementations fail to deliver any tangible results.

The common causes of CDP failure are:

1. Too hard to collect all the customer touchpoints into the platform (a problem that Littledata solves for Segment )

2. Not able to take action with the enhanced profiles, without tight integration with email marketing and personalisation tools

3. No buy-in from other areas of the company using other data stacks

No CDP has a magic bullet to resolve customer identities and build a golden profile. Your ability to make your data actionable is dependent on joining together the touchpoints on the user’s journey. This is Littledata ’s core focus, and it’s hard to do.

If your brand is trading primarily on Shopify , I know how you can avoid these traps.

You do need to get our data outside of Shopify to be future proof, but you can get many of the benefits of a CDP by using four tools: Littledata, Google Analytics, BigQuery and Klaviyo .

I’ll explain why just those four, but first of all some more context on CDPs.

What does a CDP do?

There’s no clear consensus on what a CDP must do to earn the name. For me there are four core functions:

1. Match incoming data points to a single customer profile.

2. Build segments based on properties of the customer profile.

3. Allow the export of the segments into marketing tools.

4.? Permanently record customer interactions for future analysis

Even point #1 is contentious. For an ecommerce brand there are two different types of data: anonymous data on users browsing (which can’t be linked to a named customer) and identifiable data on customers who purchased.

Some CDPs only handle identifiable customer data, but you should also track anonymous user data so:

(a)? Events can eventually be linked to a customer profile – when the customer logs in or checks out.

(b)? Events can be used to target cookie-based audiences in marketing platforms like Google Ads or Facebook Ads.

A brand would commonly use a data warehouse (e.g. Snowflake or BigQuery) as part of stack, and for many the primary use of the CDP is the conduit to get customer event data into the warehouse.

With reverse ETL tools (e.g. Hightouch and Census) this approach is reversed, so the brand first collects data into the warehouse, and then builds audiences and exports to marketing platforms from there.

Want more like this? Follow our blog

What CDPs are Shopify brands using?

The most popular choices of CDP for Shopify Plus brands, as measured by whether they have a tracking script on the storefront, are:

·? Segment (1.2%) – Littledata connects to Segment

·? Simon Data (0.3%)

·? Adobe Experience Cloud (0.3%)

·? RudderStack (0.3%)

·? Customer.io (0.1%)

·? Treasure Data (0.1%)

·? mParticle (0.1%) *

Note: As mentioned above, many brands on mParticle and others are only capturing customer-identified events, and so we’re under-counting those that don’t have a storefront tracking script.

Only 2.5% of Shopify Plus brands have any true CDP.

I’ve left out few popular choices from that list, which have some audience building functionality but are not truly a CDP:

·? Amplitude (5%) – mainly for product analytics

·? SnowPlow (3.5%) – event data collection into a warehouse

·? LiveRamp (2%) – mainly personalisation

·? Ometria (0.3%) – mainly personalisation

Why Littledata + Google Analytics

Let’s focus on that difference between handling anonymous profiles vs identified profiles.

I don’t think these two data use cases need to be combined in one CDP.

For tracking anonymous browsing data and using that to build audiences in your major marketing platform, Google Analytics provides a powerful free stack. If your company uses Google Ads to attract customers, it’s essential that you can share cookie-based audiences directly with Google Ads – and a CDP does not provide any advantages over a direct data feed to Google.

Littledata provides a turn-key solution to get all the online customer touchpoints, including post-checkout customer events, in Google Analytics and Google Ads.

Unlike Shopify’s free integration with Google Analytics, Littledata also pulls in offline customer behaviour such as recurring orders, upsells and refunds. And as Google starts to support sharing post-checkout events linked to an identifiable customer so will we.

Littledata also works with the other major marketing channels you’d want to share a cookie-based audience with: Meta (Facebook / Instagram), TikTok and Pinterest.

Where BigQuery comes in

Google Analytics has a few big drawbacks as a data platform:

1. The detailed data is only retained for a maximum of 14 months.

2. Data cannot be exported in a raw format for processing elsewhere.

3. Personally identifiable (PII) data cannot be stored.

All these problems go away when using GA4’s free link with Google BigQuery. Event-level data storage in BigQuery is permanent, low-cost, and much less complicated to maintain than an ETL tool or CDP.

Here’s what BigQuery does on top of GA4:

1. Unlimited data storage (costing just a few dollars per month stored)

2. Own the data and export any part of it – even to another data warehouse, such as Segment

3. Join identifiable customer events onto a table of customer profile information. For example, Littledata will include a Shopify customer ID with each identifiable event, and this custom dimension can be joined onto a Shopify customer list in BigQuery via this customer ID.

Littledata (as a data pipeline) + Google Analytics (as a data platform) + BigQuery (as a warehouse) provides a powerful, no code, no nonsense solution to actioning anonymous customer profiles.

Klaviyo as a CDP

If you want to cut your reliance on paid channels then you also need a smarter way to use identified customer events with email, SMS and (for those with a native mobile app) push notifications.

Klaviyo is the leading email marketing tool, and gaining an increasing share in SMS. Pushing identifiable customer events to Klaviyo provides the fastest way to deliver smarter retention marketing.

Klaviyo improved the ability to ingest all kinds of customer data over the last year, and formally launched Klaviyo CDP.?

Klaviyo CDP allows you more control over data transformations, personalisation triggers and data warehouse syncing.

Certainly you’d need data warehouse syncing if you want to do advanced segmentation, but I’d consider Klaviyo as a whole a data platform, even if you don’t pay for the ‘CDP’ add on.

But Klaviyo out-the-box doesn’t capture all of the touchpoints you need for advanced audience building. Browsing (viewing product) adds to cart, post-purchase upsells and other granular steps would typically be missed.

And so Littledata will soon be supporting Klaviyo as a destination, including enriching customer profiles with pageviews and events that happened before they registered an email address.

When you would need a full CDP

Of course, your brand might be in the 2% that really do need a CDP.?

The brands I see getting great value from Segment are in two camps:

1. They have dozens of internal tools that need the same data feed (i.e. Segment is a data pipeline for them, rather than a true CDP).

2. They have lots of non-Shopify customer interactions, maybe from physical stores or other web properties, that need to be combined into golden profiles.

If you are a larger company with many analytics and marketing tools I’d first advocate some consolidation – while you may want to empower colleagues to use the best tool for the job, every extra tool has a cost to maintain.

Segment.com – our preferred CDP partner – is proven at scale by many enterprise brands, but it’s expensive to own and maintain. Don’t buy it for the wrong reason!

Simplify your data stack

One of the most dangerous thoughts when choosing your data stack is “my business might get really complicated in the future, so let’s future proof our stack now”.

That thought leads to an urgency to buy a CDP now, when in fact the use cases you’d get immediate value from need better data in the tools you already use (Google Analytics, Meta Ads, Klaviyo etc).

Simplifying your data stack doesn’t mean adding another layer of abstraction to it. Although adding a CDP brings future flexibility it also brings implementation costs and maintenance complexity.

In summary, my message is:

• Don’t fall for the hype about CDPs: the setup is hard, and so it’s a long road to get value
• Question why you need a CDP – is it something beyond sharing cookie-based audiences back to ad platforms, or streaming to a data warehouse?
• Use Littledata + Google Analytics + BigQuery for robust, cheap, future proof event data storage
• Use Littledata + Klaviyo for building identifiable customer profiles that you can action