Era 2 decisions are more complex than Era 1 because you're no longer just giving individuals tools. You're connecting systems, introducing new workflows, and navigating organizational dynamics that didn't exist when AI was just a chat window. The decisions that matter most here are about connectivity, governance, and keeping your team focused on outcomes instead of infrastructure.

3.1 Systems decisions

3.1.1 Current state audit

The systems audit for Era 2 starts with a question that Era 1 didn't require. Which of your existing tools can AI actually connect to?

Not every tool in your stack has an API that supports the kind of integration Era 2 requires. Some tools have robust APIs and MCP support that make connection straightforward. Others have limited APIs that only expose a fraction of the data you need. Some have no programmatic access at all. Before you start planning connected workflows, you need a clear map of what's possible with your current stack.

Start with the four foundational data sources. Your documents, your internal communication tools like Slack, your external communication tools like email, and your CRM. These are the building blocks of every connected workflow in Era 2. For each one, assess whether AI can read from it, whether AI can write back to it, and what data is actually accessible through the connection. A CRM integration that only exposes account names and deal stages is very different from one that exposes call notes, activity history, and custom fields.

Assess data quality with fresh eyes. In Era 1, bad data in a copy-paste workflow was tolerable because the human could fill in gaps from memory. In Era 2, bad data in a connected workflow gets amplified. If your CRM has inaccurate deal stages, AI will build analysis on top of those inaccuracies and present them with confidence. If email threads are missing because reps use personal accounts for some conversations, AI will have blind spots it doesn't know about. The standard for data quality rises significantly when AI is consuming data directly instead of receiving curated context from a human.

Evaluate how your security and IT teams operate. This is new territory for Era 2 and it's where many organizations stall. Connecting AI to your business systems means your security team needs to evaluate AI vendors, approve data access, and define governance policies. If your security review process takes three months per vendor, you will not move at the speed this transformation requires. Assess honestly whether your security function is structured to enable fast, informed decisions about AI tooling or whether it's going to become a bottleneck.

Get your CFO aligned early. Era 2 involves investment in new tooling and potentially in new vendor relationships. The CFO needs to understand that this is an investment phase with clear expected returns, not an open-ended technology experiment. If the finance team views every new AI tool as an incremental cost to be challenged, procurement becomes a drag on the entire initiative.

Finally, resist the urge to consolidate your tooling stack before connecting AI to it. This is a common trap. A leader looks at their scattered collection of tools and decides to clean house before building AI workflows on top of them. That consolidation project takes six months and by the time it's done, the competitive window has narrowed. Connect what you have. Consolidation can come later when you have a clearer picture of which tools are actually adding value in the connected world.

“Connect what you have. Consolidation can come later when you have a clearer picture of which tools are actually adding value in the connected world.”

3.1.2 Changes required

The foundational change in Era 2 is establishing the connections between your AI layer and your business systems. This is not a massive infrastructure project. It is a focused integration effort targeting the four data sources that matter most.

Start with documents. Your team's proposals, one-pagers, playbooks, competitive intel, and internal guides need to be accessible to AI. This usually means connecting your Google Drive or document management system so that AI can reference existing materials when producing new work. A rep asking AI to draft a proposal should get output that draws on your actual proposal templates and past examples, not generic content.

Connect your internal communication tools. Slack or Teams contain an enormous amount of institutional knowledge that currently lives in unstructured conversation threads. When AI can read internal communications, it can surface relevant context that would otherwise require a human to remember or search for manually.

Connect your external communication tools. Email is where customer relationships live in granular detail. When AI can read and send email, the entire follow-up workflow changes. The rep doesn't draft in a chat window and copy to Gmail. They draft and send from the AI layer directly.

Connect your CRM. This is the most impactful connection and also the one that requires the most care. Your CRM is the system of record for your revenue org. When AI can read from it, every workflow gets richer because it has access to deal history, contact relationships, and pipeline data. When AI can write back to it, data quality improves because updates happen as a byproduct of real work instead of requiring manual entry.

Plan for connections that break or go stale. Connected workflows are only as reliable as the integrations underneath them. A CRM connection that returns yesterday's data, an email sync that misses threads, a document link that silently fails. Any of these will produce bad output that your team acts on with confidence. Someone on your team, usually RevOps or an ops lead, needs to own monitoring connection health on a regular cadence. When a connection degrades, the team needs a known fallback and a fast path to resolution. If people discover bad data in an AI output and nobody can explain why or fix it quickly, they will abandon the connected workflow and go back to copy-paste. Trust is hard to earn and easy to lose.

The build versus buy decision becomes real at Era 2. Some members of your team, usually your most technical and motivated people, will start building custom integrations and workflows. This is a double-edged dynamic. Their initiative is valuable because it validates demand and surfaces what connected workflows should look like. But if left unchecked, you end up with fragile, undocumented custom builds maintained by people whose actual job is selling or managing accounts. Establish a clear framework early. Use building as a way to validate the need. Then go buy the mature third-party solution that does the same thing with proper support and maintenance. Do not let your best revenue people become part-time infrastructure engineers.

“Use building as a way to validate the need. Then go buy the mature third-party solution that does the same thing with proper support and maintenance.”

3.1.3 Sequencing

First, connect documents. This is the lowest-risk, highest-immediate-value connection. Your AI workflows get significantly better when they can reference your actual templates, playbooks, and historical materials. The integration is usually straightforward and doesn't involve sensitive transactional data.

Second, connect internal communications. Slack or Teams integration gives AI access to the institutional knowledge that lives in conversation threads. This connection adds context depth to every workflow.

Third, connect external communications. Email integration is where the workflow shift becomes tangible for your team. The ability to draft and send from the AI layer eliminates the copy-paste step that defined Era 1 and makes the connected workflow feel real to individual contributors.

Fourth, connect your CRM. This is the most complex connection and the one with the most governance implications. By sequencing it after the other three, you give your security and IT teams time to build comfort with AI data access through lower-stakes integrations first. You also give your team time to build fluency with connected workflows before adding the system of record to the mix.

Run each connection as a pilot with a small team before rolling it out broadly. Validate that the integration works reliably, that the data quality is sufficient, and that the workflows built on top of it produce the expected value. A pilot with five to ten users for two weeks is enough to surface the issues that a full rollout would amplify.

The full systems sequence for Era 2 takes two to three months from first connection to full deployment across the team. The connections themselves are fast. The governance, testing, and rollout take the time.

3.2 Technology decisions

3.2.1 Current state audit

The technology audit for Era 2 builds directly on the systems audit. You know which systems can be connected. Now you need to evaluate the AI layer that will connect to them.

Start with your core AI platform. In Era 1, you picked Claude or ChatGPT and got everyone licensed. In Era 2, the question shifts from "which AI tool" to "which AI platform supports the connected workflows we need." Not every AI tool can connect to external systems through MCP or equivalent protocols. Evaluate whether your current AI tool supports the integrations you identified in the systems audit. If it doesn't, you may need to switch or add a platform that does.

Assess what your team has already built. This is the shadow IT problem of Era 2. Your most motivated people have likely started connecting tools on their own. They've built Zapier automations, written scripts that pipe data between systems, or configured AI features inside individual tools. Map all of it. Some of these builds are valuable prototypes that show you what connected workflows should look like. Others are fragile, insecure, or duplicative. You need the full picture before you decide what to formalize, what to replace, and what to shut down.

Evaluate the security and compliance posture of every AI tool that touches your data. In Era 1, the risk surface was limited to what individuals pasted into chat windows. In Era 2, AI has direct access to your CRM, your email, your customer data. The security bar rises significantly. Every tool that connects to your systems needs enterprise-grade data handling, clear data processing agreements, and contractual protections. If a tool your team loves doesn't meet the security bar, the answer is to find one that does, not to lower the bar.

Look at the platform versus point solution landscape. You will be tempted to buy multiple specialized AI tools, one for sales engagement, one for CS health scoring, one for email automation, one for pipeline analytics. Each will promise deep functionality in its niche. The risk is that you end up with a collection of disconnected AI tools that recreate the same fragmentation you had before AI. Evaluate whether a platform approach that connects a single AI layer to all your systems might serve you better than a stack of point solutions that each connect to one system.

3.2.2 Changes required

The central technology change in Era 2 is establishing the context layer. This is the infrastructure that gives AI access to your business data across systems.

For most organizations, this means configuring your core AI platform to connect to your business tools through MCP or similar protocols. The practical work involves setting up connections, defining what data AI can access, establishing permissions, and testing that the workflows produce accurate and useful results.

Address the platform versus point solution question deliberately. A single AI platform connected to all your systems creates a unified context layer where every workflow benefits from the full picture. A collection of point solutions creates depth in individual workflows but no cross-system intelligence. The right answer depends on your team's specific needs, but lean toward the platform approach where possible. The value of Era 2 comes from connected context, and that value diminishes when context is fragmented across multiple tools.

Formalize the useful builds your team has already created and sunset the ones that are fragile or redundant. For the automations and integrations your people built that actually work well, evaluate whether to migrate them to a supported platform or to use them as specifications for what a third-party tool should do. For the ones that are held together with tape, replace them with proper solutions before they break in production.

Set up governance for AI tool access. In Era 1, governance was a one-page data guideline. In Era 2, you need a clear policy on which tools are approved to connect to which systems, who can set up new connections, and how new tools get evaluated and approved. This doesn't need to be bureaucratic. But it needs to exist because the risk surface is larger and the consequences of a bad connection are more significant.

3.2.3 Sequencing

First, validate that your core AI platform supports the connections you need. If it does, configure the first connection from your systems sequence. If it doesn't, evaluate alternatives now rather than halfway through the rollout.

Second, audit and formalize shadow AI builds. Surface everything your team has already created. Decide what to keep, what to migrate, and what to sunset. Do this before you roll out official connected workflows because you don't want competing systems running in parallel.

Third, deploy connected workflows through the same sequence as your systems connections. Documents first, then internal comms, then email, then CRM. Each connection gets its own validation cycle before the next one begins.

Fourth, establish the governance framework. Approval process for new AI tools. Clear ownership of connections and integrations. Regular review of what's deployed, what's being used, and what's not. This framework should be lightweight enough that it doesn't slow adoption and robust enough that your security team can trust it.

The technology sequence runs in parallel with the systems sequence. They are the same work viewed from two angles. Total timeline is the same two to three months.

3.3 People and org decisions

3.3.1 Current state audit

The people audit at Era 2 is fundamentally different from Era 1. In Era 1, you were assessing who was using AI and who wasn't. In Era 2, you're assessing who is making the transition from individual AI user to systems thinker, and who is getting stuck or going sideways.

Start with your Era 1 power users. These are the people who figured out AI early and got the most value from it. Some of them will naturally evolve into Era 2. They'll start thinking about how to connect systems, how to build repeatable workflows, how to make their individual productivity gains available to the team. These are your Era 2 leaders. Others will get stuck in copy-paste mode because it works well enough for them and the incentive to change isn't strong enough. Identify who is evolving and who is plateauing.

Watch for the infrastructure rabbit hole. This is when someone whose job is to sell or manage accounts quietly becomes a part-time systems engineer instead. Some of your most motivated people will start spending significant time building AI workflows, creating integrations, and engineering systems. This looks like initiative and innovation. Sometimes it is. But sometimes it's a talented rep or CSM who has effectively stopped doing their primary job while their pipeline or their book of business suffers. The line between productive experimentation and the infrastructure rabbit hole is whether the person is still hitting their number. Assess this honestly.

Evaluate your management layer with a specific focus. Can your frontline managers tell the difference between someone who is productively building connected workflows and someone who is disappearing into engineering work? Can they review AI-generated outputs and provide useful feedback? Do they understand what connected workflows look like well enough to coach their teams on using them? If the answer to any of these is no, your management layer needs investment before the team will successfully make the Era 2 transition.

Look at the people who serve as connective tissue in your org. The RevOps analysts, the CS operations people, the deal desk coordinators. These are the people who currently spend much of their time moving data between systems, building reports from multiple sources, and maintaining the operational infrastructure that keeps the revenue org running. Connected AI workflows will change their roles significantly. Some of the work they do today, the manual data movement and report assembly, will be handled by AI. Assess how they're responding to that shift. The ones who see it as an opportunity to focus on higher-value analysis and strategy are your allies. The ones who see it as a threat to their role need a clear picture of what their job becomes.

Assess the organizational readiness for procurement speed. Era 2 involves buying and deploying new tools faster than most organizations are comfortable with. Your security team needs to evaluate vendors. Your IT team needs to support integrations. Your finance team needs to approve spending. If any of these functions are slow, risk-averse, or understaffed for the pace you need, the people constraint will be more limiting than the technology constraint.

3.3.2 Changes required

The most important people change at Era 2 is establishing clear boundaries around build versus buy so your best people don't become infrastructure engineers.

Your most talented, most motivated team members will want to build. They see the potential of connected workflows and they have the skills to start wiring things together. That energy is valuable. It validates demand and shows you what workflows matter. But if you let it run without guardrails, you end up with critical workflows maintained by someone whose day job is closing deals. When they leave, the workflow breaks. When they go on vacation, nobody knows how it works.

The framework is simple. Let people build to validate need. When a workflow proves its value, replace the custom build with a bought solution that comes with support, documentation, and maintenance. Use the homegrown version as a specification for what the bought solution needs to do. This keeps the innovation energy flowing while preventing the accumulation of unsupported technical debt inside your revenue org.

Managers need to develop a new skill at Era 2. They need to recognize the difference between productive systems work and unproductive distraction. A rep who spends two hours building a connected workflow that saves the entire team ten hours a week is creating real value. A rep who spends twenty hours building a complex automation that saves them thirty minutes a day is probably avoiding their pipeline. The distinction requires judgment and context. Managers who aren't using connected AI workflows themselves can't make this distinction. Manager enablement is a prerequisite for team enablement.

“A manager who isn’t using connected AI workflows themselves can’t make this distinction.”

The connective tissue roles need a clear path forward. The people who currently spend their days moving data between systems need to understand what their job looks like when AI handles the data movement. The answer is that they become the people who design, maintain, and optimize the connected workflows. They go from executing repetitive data tasks to owning the systems that do it. Frame this transition clearly and early. If these people feel threatened, they will resist the change. If they see a path to a more strategic and more interesting role, they'll help drive it.

Reorient your security and procurement functions as enablers. This is not a suggestion. It's a requirement. If your security team sees its job as saying no to new tools, Era 2 will stall. Security needs a mandate to evaluate AI tools proactively, set clear criteria for approval, and move at a pace that matches the business need. This might mean adding headcount to the security team or bringing in outside expertise to handle the evaluation load. The cost of a fast, competent security function is trivial compared to the cost of losing three months to vendor review delays.

3.3.3 Sequencing

First, set the build versus buy framework. Before you connect any systems, make it clear that individual builds are experiments, not permanent infrastructure. Celebrate the experimentation. Set the expectation that successful experiments get replaced with bought solutions. This framing prevents the infrastructure rabbit hole before it starts.

Second, enable your managers. Get them on connected workflows before their teams. Managers need firsthand experience with the new way of working before they can coach others on it. A manager who has used a connected deal prep workflow can spot the difference between productive and unproductive systems work. A manager who hasn't is guessing.

Third, run the first connected workflows with a pilot team. Pick five to ten people who are already leaning into Era 2 behaviors. Let them test the connected workflows, surface the problems, and refine the processes. Their feedback shapes what the broader rollout looks like.

Fourth, expand to the full team with clear enablement. Show people how the connected workflows work. Demonstrate the value with real examples from the pilot team. Make the new workflow easier than the old one. If the connected workflow is harder than copy-paste, people will revert.

Fifth, address the connective tissue roles. Once connected workflows are running, sit down with the people whose manual data work is being replaced. Show them the new version of their role. Give them ownership of the workflow systems. Invest in the skills they need to manage and optimize those systems.

The people timeline for Era 2 is longer than the technology timeline. The connections can be up in two to three months. The full behavior change across the team takes four to six months. The connective tissue role transitions can take six to nine months. Plan for the longest timeline, not the shortest.