Tableau Acquisition: How We Re-Prioritized CRMA Roadmaps for 2019-2020
Yesterday Salesforce announced it is acquiring Tableau for approximately $15.7 billion. That is, as of this writing, the largest acquisition in Salesforce's history. Our phones started ringing before the press release had finished loading.
We want to be direct about what we know, what we do not know, and how we are advising clients right now — because the honest answer is that nobody outside of Salesforce's boardroom has a clear picture of what this means for Einstein Analytics over the next eighteen to thirty-six months. What we do have is eleven years of watching platform acquisitions ripple through enterprise implementations, and some grounded thinking about how to navigate the uncertainty without freezing roadmaps or making reactive decisions either way.
What We Actually Know as of June 11, 2019
The deal is announced. It has not closed. Integration planning, if it exists in documented form, has not been shared with the partner ecosystem. Einstein Analytics is still the product. Tableau is still Tableau. Nothing in either product has changed overnight.
That said, the strategic signal is loud and worth taking seriously. Salesforce has been building Einstein Analytics as its native analytics layer — embedded in the platform, governed by Salesforce data models, natively integrated with CRM data, and increasingly tied into Einstein AI capabilities. Tableau has roughly 86,000 customers and a market position built on visual analytics and broad data connectivity that goes well beyond CRM data. These are not redundant products. They are also not seamlessly complementary ones. There is real overlap in dashboarding and exploration use cases, and that overlap is going to create genuine confusion for buyers and existing customers alike.
We are not going to speculate about product roadmap convergence because we have no basis to do so responsibly. What we will say is that the questions our clients are asking right now are reasonable, and deserve honest answers rather than reassurance.
The Questions Clients Are Asking
Within the first twenty-four hours we fielded variations of the same four questions from clients at different stages of Einstein Analytics adoption:
"Should we pause our Einstein Analytics implementation?" No, with caveats. If you are mid-implementation and have already invested in data modeling, recipe development, and dashboard buildout inside Einstein Analytics, stopping now creates costs without creating clarity. The platform is not going away. Salesforce has too much existing Einstein Analytics revenue and too many commitments to enterprise customers to deprecate it on any near-term timeline.
"Should we evaluate Tableau instead of continuing with Einstein Analytics?" This deserves a more nuanced answer. If your analytics use case is primarily about Salesforce CRM data — pipeline analysis, service metrics, marketing attribution against Salesforce objects — Einstein Analytics is still the more natural fit today. If your use case involves substantial external data sources, enterprise-wide BI across non-Salesforce systems, or a data team that already has Tableau expertise, you now have a stronger rationale to evaluate Tableau seriously and not feel like you are choosing against the Salesforce ecosystem.
"Will our Einstein Analytics investment become obsolete?" We do not believe so, but we also cannot promise a specific roadmap. What we can say is that Salesforce has historically maintained acquired products for long periods post-acquisition, and the Einstein Analytics user base is not small. The more useful question is whether the capabilities you need today and over the next twelve to eighteen months are available in Einstein Analytics. For most of our CRM-centric clients, the answer remains yes.
"Is this a sign that Einstein Analytics was not working?" We actually do not read it that way. The acquisition reads more like Salesforce recognizing that the enterprise analytics market is bigger than CRM analytics, and that building that breadth organically would take years. Einstein Analytics was building well within its lane. This acquisition expands the lane.
How Our Practice Is Responding
We spent most of yesterday afternoon in an internal conversation about how to adjust our advisory stance. Here is where we landed.
We Are Not Changing Our Einstein Analytics Build Recommendations for Existing Clients
For clients who are actively building on Einstein Analytics — dataflows, recipes, SAQL queries, dashboard components, embedded analytics in Lightning pages — we are not recommending any change in approach. The architectural decisions that make an Einstein Analytics implementation solid are the same ones that would make any future migration or integration tractable: clean data models, documented dataflow logic, reusable dashboard components, proper permission set architecture.
Implementations that were already a mess do not become more manageable by waiting to see what Salesforce announces in six months. Implementations that are being built thoughtfully will be easier to adapt regardless of what platform direction emerges.
We Are Adding a Tableau Validation Track to New Engagement Scoping
For clients who are still in the evaluation and scoping phase — those who have not yet made a significant Einstein Analytics investment — we are now explicitly including a Tableau fit assessment in our scoping process. This is not us hedging against Einstein Analytics. It is us doing our job, which is recommending the right tool for the specific use case.
The honest reality is that before this announcement, recommending Tableau to a Salesforce-centric organization felt like recommending something slightly outside the ecosystem, with the friction that implies. That friction just changed meaningfully. If a client has a substantial external data story and an internal data team that can manage Tableau Server or Tableau Online, we want to evaluate that path seriously rather than default to Einstein Analytics because it is the familiar choice.
We Are Building Internal Knowledge on Tableau's Data Connectivity Model
Several of our senior consultants are spinning up on Tableau's data source and extract model, specifically to understand how Salesforce data behaves when accessed through Tableau connectors versus natively in Einstein Analytics. We expect this to be a recurring client question and we want to give informed answers rather than theoretical ones.
Early observation: the integration between Tableau and Salesforce data today is functional but not as seamless as native Einstein Analytics access to Salesforce objects. That gap may close significantly post-acquisition, but it is a real consideration today for clients evaluating a Tableau-first path.
The Honest Tradeoffs Right Now
We want to put something plainly that we sometimes see glossed over in analyst commentary.
Einstein Analytics is a strong product for what it is designed to do. It is not a general-purpose enterprise BI platform. It has real constraints around data volume in some configurations, it has a learning curve on SAQL for complex calculations, and the recipe-based data preparation layer, while improved, is not as mature as dedicated ETL tooling. These were real tradeoffs before yesterday, and they remain real tradeoffs today.
Tableau has enormous depth in visual analytics and broad data connectivity. It also requires infrastructure decisions — Tableau Server versus Tableau Online, licensing model complexity, a data preparation and governance story that is less native to Salesforce's security model — that add friction for organizations that are primarily Salesforce shops.
The acquisition does not resolve these tradeoffs. What it does is put both products on a potential convergence path that neither product is on today. How long that convergence takes, and what form it takes, is genuinely unknown. We are not going to pretend otherwise to avoid making clients anxious.
What We Are Watching For
A few things will help us update our thinking significantly when they become clearer:
Integration roadmap communication from Salesforce. Even high-level signals about whether these products are being positioned as complementary, converging, or eventually unified will change our advisory posture. We expect something at Dreamforce 2019 and we will be paying close attention.
Licensing and packaging changes. Einstein Analytics is currently sold as part of the Salesforce platform licensing structure. Tableau has its own licensing model. How Salesforce harmonizes this — or whether it does — will affect recommendations for net-new buyers significantly.
What happens to Einstein Analytics product development velocity. One risk in a large acquisition is that internal product teams get pulled into integration planning and slow down feature shipping. We will be watching the Einstein Analytics release notes closely through the remainder of 2019.
Early signals from Tableau's data team on Salesforce connectivity. If Tableau invests in deeper native connectors to Salesforce's data model — not just SOQL-based connectivity but something that understands Salesforce's metadata model more natively — the evaluation calculus changes. We are watching for technical previews.
Our Actual Recommendation as of Today
If you have an active Einstein Analytics implementation: keep building, build it well, and do not make reactive architectural decisions based on an announcement that has not yet translated into any product changes.
If you are evaluating analytics on Salesforce for the first time: this is the right moment to do a genuine two-path evaluation rather than defaulting to Einstein Analytics as the only serious option. The ecosystem context just shifted. Use that context.
If you are a Tableau customer wondering about your relationship with Salesforce: nothing practical has changed yet, but the direction of travel is toward tighter integration. That is probably good news for you if you have been wanting better Salesforce connectivity. How quickly it materializes is the open question.
We have been through enough platform acquisitions to know that the first ninety days are high on anxiety and low on actionable information. The best thing most organizations can do right now is maintain technical discipline on whatever they are building, document their current architecture thoroughly so they can make informed decisions when better information arrives, and avoid letting uncertainty become a reason to defer investment decisions that were already justified.
The analytics layer inside Salesforce-centric enterprises is a genuinely important problem. Both of these products are real attempts to solve parts of it. The fact that Salesforce now owns both of them does not simplify the decision landscape today, but it does suggest there is a longer-term story here that is worth staying engaged with.
We will continue publishing our observations as this develops. If you are working through platform decisions right now and want to talk through specifics, our contact information is below.
This article reflects our perspective as of June 11, 2019. Platform recommendations in this space are evolving rapidly and we will update our guidance as more information becomes available.