TSIA Envision October 2025 Preview: We Answer the Top AI Questions Tech Services Leaders Are Asking

TSIA Envision 2025 is right around the corner, and tech services leaders everywhere are asking the same urgent question: How do we prepare our people, our services, and our business models for the AI era without losing relevance?

As we get ready to meet you in Las Vegas, the Seertech team has been thinking deeply about these conversations. To get into the right mindset ahead of the conference, we picked our Chief Strategy Officer, Scott Mahoney’s brain. Scott has spent the last 25+ years helping Fortune 500 organizations — from Microsoft and IBM to Cisco and GE — transform learning from a checkbox activity into a true business growth engine. Few people understand the intersection of AI, workforce capability, and business value more than Scott.

Like TSIA poses on the Envision conference’s website, we asked Scott the 7 biggest questions every tech services professional is asking about AI…read his take below!

How to find us at TSIA Envision 2025

If you’re attending the conference, we’d love to connect in person! You have some options:

• Visit us at Booth #523
• Book a strategy session with our Customer & Partner Education advisors, Emily and Kealan
• Join the Interactive AI Strategy Workshop led by TSIA thought leader Janice Lee – From Benchmark to Blueprint: An AI Strategy Workshop for ES Leaders
• Or grab a seat at our exclusive dinner + networking meetup

👉 All the details and session booking links are here.

Question 1: Will AI make my role obsolete?

Scott’s take: AI rarely eliminates whole roles; it re-slices task bundles. Jobs shift from doing work to designing, supervising, and compounding work through AI, preparing the work to move up-market (up-value). Successfully navigating AI introduction is defining your value proposition and moving it up the value tree, using AI as the tool to scale your value.

Question 2: How fast do I reskill — and in what areas?

Scott’s take: I take this question 2 ways: How fast do I reskill in AI and in what areas; and How fast do I reskill in areas AI is unlikely (in the short-medium term) to replace?

Let’s look at answers to both of these questions:

1.How fast do I reskill in AI and in what areas

If we are talking about AI reskill, aim for T-shaped AI fluency in 90 days: create broad literacy + a deep spike aligned to your role. T-shaped means broad, cross-functional literacy (the “horizontal bar”) + one deep, role-specific capability (the “vertical spike”). It lets everyone work safely and productively with AI, while also allowing you to become the go-to person for a specific, high-value workflow.

2.How fast do I reskill in areas AI is unlikely (in the short-medium term) to replace

If you are talking about reskill into areas not likely to be affected by AI, then you will need to ‘aim to be safe’ as quickly as possible, as decisions are being made now on role alignment and need. For example, I would aim to be “safe” in ~90 days via a combination of understanding the impacts and use of AI, demonstrate your literacy and competency in AI to remain relevant, but aim to shift into shaping the inputs and outputs of AI usage via a focus on durable skills development:

Question 3: Will leadership roles shrink? Org design for AI-native teams

Scott’s take: Leadership in general doesn’t shrink; it shifts to platform thinking, high value orchestration, and evaluation culture. This may result in some middle management job losses that are focused on pure orchestration (and where the incumbents cannot pivot to the new mode of thinking and execution), but leadership roles in general will reimagine and redeploy that are focused on determining and curating, measuring and optimizing the effective use of AI toolsets. If anything, these roles will become even more in demand as organizations understand both the capabilities (and limitations) of AI tools

Question 4: How do I prove my value? Metrics, KPIs, and attribution

Scott’s take: You need to prove your value through proving your role (and output metrics) in terms of business outcomes, not just activity. Tie everything you do, every workflow you own or contribute to, to business cost, quality, growth, or risk metrics. Business impact and value realization is now THE benchmark you will be evaluated against. If you don’t think this way, you will be sidelined and replaced.

Question 5: Will compensation models change? Consumption & outcome pricing

Scott’s take: Let’s take this question and split it, as these are separate but somewhat related topics

Will Compensation Models Change?

Yes. Comp should reward value creation and safe leverage of AI, not raw output. In other words the measure of KPI will be based more on the QUALITY of business impacts, not just the amount of work done. At the same time, organizations will be forced to de-silo to force optimization of outputs (that is a whole other area AI will have an impact on), which means there will be more need for, and reward for collaboration to broaden business impact vs local optimization.

Compensation Levers by Function (TSIA members)

• CSM: Base + NRR/GRR + adoption milestones for AI features; team modifier for deflection quality.
• Support: Quality-weighted throughput (FCR, TTR) + customer-rated outcomes; guardrail: no reward for unsafe automation.
• PS/Services: Hybrid fee: base + success component tied to outcomes (time-to-value, KPI lift) with caps and clear SOW scopes.
• Sales: Credit AI-assisted pipeline with policy-defined attribution; avoid overpaying for AI-generated content volume.
• Leaders: Portfolio ROI and risk-adjusted outcomes; cost-to-serve targets.

Guardrails Needed: No pay for hallucinated output; quality gates before crediting; shared incentives to avoid local optimization.
How will Consumption & Outcome Pricing Change?
AI is reshaping consumption & outcome pricing for tech + services. The trick will be how to do it without blowing up margin. AI shifts pricing from time & seats to usage & verified outcomes. The trick is to meter the right unit of value, quality-adjust it, and wrap it with cost/safety guardrails.

Key directional changes:

• Bill for outcomes, not tokens. Make quality the meter and tokens the guardrail.
• Small risk-share, big trust. Collars + neutrality bands keep it fair.
• Governance is a product. Change windows, canaries, and version pinning belong in the contract.

This will also result in shifts in contracting – especially around evaluation of success criteria and tying this to accelerators and roll-backs, as well as ‘back-out’ clauses based on performance metrics. It will also require AI assisted FinOps to ensure that margins are measured dynamically based on output performance vs cost and commitments.

Question 6: Will domain expertise still matter? Building a moat

Scott’s take: Domain depth becomes more valuable as models commoditize. Moat = proprietary knowledge × evaluation sets × relationships. Domain expertise is also absolutely essential to get the most out of AI tools, as they are fundamentally pattern matching toolsets that require strong guidance using domain expertise to recognize, avoid or address hallucinations / false positives from AI. AI is a tool – it doesn’t replace the worker’s ‘brain’ but is a force multiplier in output efficiency – if used in the right way. That is up to you.

Turn Expertise into Leverage

• Capture: Curate reference designs, policies, and decision trees; normalize docs for retrieval.
• Structure: Light ontologies/taxonomies; document schemas; entity dictionaries.
• Teach: Build golden Q&A and scenario rubrics; run red-teaming for edge cases.
• Operate: Owners for content freshness SLAs; change logs; semantic diff reviews.

Question 7: Am I at the right company for the AI era? Incumbent advantage vs. startup agility

Scott’s take: You need to pick companies in environments with a data advantage, shipping cadence, strong leadership with an innovation and evaluation culture. In the Tech Services industry, these are key benchmarks for competitive advantage whether you are in an established business or startup.