Why a Bundled Analytics, Testing, and Personalization Solution May be Right for your Business

Is 'Composable' a binary or a spectrum? Many platform solutions perform quite well as part of a larger stack.

Last post went deep on composability and the details and particular benefits of taking a composable approach to your overall architecture and technology stack. I see a lot of benefits from the composable approach, especially for brands that have the commitment, talent, and bandwidth to do the heavy lifting of tying everything together properly with sound integrations and front-end development.

That said, a bundled solution is often the right decision for many brands, especially if you tend to run lean on the solution design, IT, and development side of the house.

This post highlights the following key takeaways (in case you’re looking for the too-long-didn’t-read):

• Bundles solutions that extend strong core capabilities and properly unify additional capabilities within a single UI have a ton of value to marketers, analysts, and product owners by enabling needed capabilities with less implementation and integration overhead

• The composable/monolith binary is not as binary as people think: our example case of Amplitude helps demonstrate this by highlighting both the benefits of the unified capabilities within the Amplitude platform but its flexibility and extensibility within a larger, ‘composed’ technology stack

What's in a Bundle?

For many brands, building a properly composable solution along with the work to tie everything together is neither feasible nor desirable. The considerable benefits of a composable stack are often outweighed by the advantages of time-to-value, onboarding time, diminished development needs, and potential simplicity of a bundled solution or suite (or monolithic solution as advocates of composability might call it). To get full value out of a platform solution, there is still plenty of integration, data infrastructure, and implementation that may need to happen, but it can be considerably less than the integration work needed to tie a number of best-in-breed components together in a properly composable architecture.

First and foremost, let's clarify the range of platforms and solutions that satisfy the bundle definition. There clearly is not a one size fits all definition for this. On the SMB/mid-tier side of things, there are suites that tackle a very particular set of needs related to Martech and Adtech without offering a fully baked all-in-one ecosystem you buy into. Similarly, there are platforms that started out addressing a single capability and have slowly expanded to include a bundle of related capabilities: Klaviyo is a perfect example of this in the customer engagement arena. They built a great business serving customer engagement needs with a strong focus on Shopify ecommerce stores, and they have expanded their capabilities to include multiple engagement channels, robust journeys, and now have many CPD-light capabilities as part of their platform.

A standout in the bundled solutions arena with a clear enterprise focus with mid-tier appeal and accessibility is Amplitude. From their core offering, a proper product analytics solution for mobile apps and web sites, Amplitude has invested in expansion of their core digital capabilities over the last five years (and they continue to do so). At this point, Amplitude is clearly a bundled platform (though you can still utilize their product analytics standalone).

Amplitude built robust capabilities for launching tests/experiments directly via the Amplitude platform, personalization, session replay, feature flagging, user surveys and guides, direct integrations with numerous downstream activation channels, warehouse native capabilities, and a host of other powerful capabilities. The recent addition of a strong agentic component to Amplitude shows the company’s commitment to continuous investment in their platform. This is not your run-of-the-mill box checking exercise so that they can say (along with everyone else): “our platform does AI.” The agentic capabilities rolling out allow teams to hand over some of the more tedious, repetitive, and time consuming aspects of activating on behavioral data by specifying desired outcomes, improvements, or end states while utilizing Amplitude’s AI agents to perform the set up and execution autonomously. This frees up your experts to focus on high-complexity, high-value human workstreams while simultaneously building scale and velocity into your testing and improvement programs.

To not oversimplify, it is important to understand that Amplitude requires implementation work as any proper analytics platform does. Yes, you can implement a single line of code on your website to enable automatic data collection (the mobile app SDKs also support some autocapture measurement) similar to Google Analytics 4 and other solutions, but to meet the needs of most brands it is important to collect additional custom events and context variables to enrich your data and support segmentation and other analysis. I use Amplitude to expand on properly bundled platform solutions below, and we will get into how Amplitude’s integrated capabilities do set it apart from the crowd in many respects.

On the paid media side, Google continues to position its Google Marketing Platform suite, including Display & Video 360, Campaign Manager 360, Search Ads 360, and Google Analytics 4 as a feature-rich toolset for multi-moment digital marketing programs. In its ideal conception, marketing teams can use this suite to integrate media measurement, behavioral analytics & audience segmentation, ad serving & attribution, and programmatic media buying in a neatly packaged and integrated solution.

To get the most out of the Google Marketing Platform stack, it is no secret that a good deal of data integration and development work is necessary in order to support value-based bidding scenarios, custom bidding scripts for programmatic, offline conversion ingestion, advanced media analytics using data transfer files in Google Cloud, etc. GMP supports activation and optimization within the Google ecosystem very well. Opinions are mixed, at best, about GMP's ability to deliver similar results for the open web.

For larger agencies and the big media agency holding companies, there are competing solutions from MediaOcean and others. There are also strong standalone solutions for particular capabilities, especially Demand-Side Platforms (DSPs) and attribution tools.

Moving into the full enterprise waters, there are a handful of platforms that both position themselves and offer (at least in theory) fully bundled ecosystems for brands: Adobe and Salesforce are top of mind in this category. These 'fully bundled' options offer everything from content management, customer relationship management (CRM), user experience, commerce engines, service, customer engagement, analytics, paid media management, and customer data platform capabilities, as well as a number of other core digital capabilities. Depending how you look at these enterprise suites, they are either massive monolithic solutions or they are aspiring composable stacks with the ability to pick and choose some capabilities from the core stack and then integrate and layer in other solutions, as needed. Both Salesforce and Adobe have grown capabilities through aggressive acquisition activity over the past decade and more, and some of the key areas where each suite shines are reflected in the core areas of the platform (or what you might call the 'true' DNA of each company).

How far you are willing to stretch to call Adobe and Salesforce platforms composable in the proper sense of the word probably has a lot to do with which features of each suite you are using. There is a strong case to be made, for example, that Adobe Commerce can be the commerce engine component of a composable stack as it was formerly the Magento platform. The open source, headless option can be flexibly integrated in a number of ways with other compatible solutions. How well this type of approach works out has more to do with how well your team designs, integrates, and deploys your solution than it does with any inherent strengths or weaknesses of Adobe Commerce.

On the Salesforce front, there is a really good reason why Salesforce has such an extensive network of Salesforce professionals consulting on everything from the capabilities themselves to designing a specific solution to implementing and maintaining these complex environments. Salesforce's offering at this point is not just extensive and expensive but also complex and requires thoughtful planning to implement and utilize its capabilities properly. Once you move beyond a simple instance of a Salesforce CRM to truly incorporate the enterprise suite of capabilities (not to mention the hundreds of integrated third-party marketplace solutions) you are in territory that has a lot of elements of both a monolithic suite and a composable solution. And however you decide to label the large enterprise suites, some of the distinctions between composable and monolithic start to break down when you get sufficiently into enterprise solution territory. The fact is that Adobe, Salesforce, and any other proper enterprise suite always requires significant customization, integration, front-end development, and very rarely delivers every capability it has in its quiver well enough that other third-party solutions are unnecessary.

Hopefully this discussion highlights that when we talk about fully-composed suites or solutions, we may be speaking about a number of different types of platforms and capabilities. A useful summary would be to define bundled or composed solutions as platforms or suites that offer multiple core capabilities within a single UI (or at least a coupled and integrated UI) that could otherwise be licensed and implemented separately and integrated together as part of a larger ecosystem or stack.

Why Should You Care About Bundled Solutions?

Bundled solutions offer a lot of compelling advantages, especially from the industry players that are doing it right and investing in powerful capabilities that rely on shared data backbones with seamless integrations that help you generate better insights, improve performance, activate on your data, and intervene directly in user experience.

You don't need to procure another license and implement another tool with really good bundled platforms. That's not to say there are no cost implications. Of course, increased utilization of platform features will increase your overall utilization level of that platform (and corresponding fees). Let's break that down using Amplitude as our example because they have genuinely built a strong suite of bundled capabilities well worth consideration for many brands.

• implement their core product analytics solution under a paid license to support my data volume and needs (you start to incur monthly costs based on your data collection volume, as expected)

• start utilizing their experimentation features because I am convinced that being able to target behavioral segments and generate user experience tests directly from Amplitude will make my product team's life easier and accelerate our test & learn capabilities (overall platform utilization and data volumes and my Amplitude fees increase, as expected)

• Duly impressed by Amplitude's integrated experimentation functionality, you start using their user guides and user surveys capabilities, too. Now you are intervening directly in the user experience and asking key questions of users to better understand their experience and potential friction points. Not just that, but you launch and serve guides and nudges to help move them through those friction points or onboarding processes that tend to stall adoption or lead to abandoned conversion paths (this further increases utilization, data volumes, and Amplitude fees, as expected).

Yes, as you use an increasing number of integrated capabilities, you will pay more for a platform such as Amplitude. But the basis of comparison is not your base Amplitude license costs. If you need the capabilities offered, if you sourced them separately from standalone platforms, you'd be paying for an additional testing tool (Optimizely, VWO, StatSig, etc.) a survey tool, possibly a customer experience platform, and you may find you are layering in a number of other platforms and middleware to handle the capabilities bundled in Amplitude.

This raises two questions: how significant are the benefits of having product analytics, experimentation, and survey and guide capabilities that drive various types of optimization, user experience, and qualitative research needs within an integrated toolset compared to the headache of sourcing these capabilities separately, paying these separate license costs, and trying to ensure that all of the component parts are well-integrated, interoperable, and actually play nice together?

Amplitude’s platform raises another question on this composable/composed divide: i.e. is there really such a firm chasm between composability and bundled solutions in the first place. Amplitude’s data backbone and supported integrations are such that it can be (and is) flexibly integrated and paired with numerous other technologies, both upstream and downstream. For example, you can fairly seamlessly use a robust data collection schema of Twilio Segment and send the data directly into Amplitude. Downstream, you can pair with a customer engagement platform (CEP) such as Braze for customer activation via email, sms, and push. This may not meet the full ‘orthodox’ definition of composability (e.g. something the like the MACH Alliance standard), but we’re also not a priesthood of true believers in digital. Composability is much more of a spectrum than a dichotomy, and bundled platforms with a solid data backbone with numerous integration options upstream and downstream help to drive home the value of bundled solutions and composable possibilities.

In my experience, the benefits of having closely aligned capabilities that rely heavily on shared data infrastructure and identity spaces coupled or bundled generally creates better audience fidelity across those capabilities, generally ensures better delivery of the expected user experiences and audience activations, and makes performance reporting much more seamless. It's for these same reasons that Google's Marketing Platform stack delivers much better results within the Google universe than it does within the broader paid media landscape. It is to be expected.

Want to have a conversation about this? I invite you to reach out and connect.

Get In Touch