Go-To-Market Transformation: Improved Marketing, Sales, And Customer Service

Evidence and data. There are many use cases for go-to-market transformation, and while some are already implementing, others will manifest over time.⁶ (See figure below.) Interviewees said that by using AI agents to deliver actionable, contextual insights, their organizations elevated their marketing, sales, and customer service functions, creating more meaningful and relevant interactions rather than simply increasing scale. AI agents also supported and enhanced human teams, equipping them with new ways to engage buyers and drive successful customer outcomes. Both business-to-employee (e.g., agents that salespeople use) and business-to-customer (e.g., agents that interact directly with prospects and customers) use cases increased revenue and customer satisfaction.

Interviewees and survey respondents said their organizations increased revenue from improvements in their marketing, sales and customer-service functions. Some of the improved metrics include the number of qualified leads, win rates, and customer retention.

• The managing director of strategic supplier relationships and AI transformation at a media and entertainment company described agents assisting with assessing and responding to RFPs and said for more than 80% of the RFPs, the time to respond was reduced from weeks to a day — and with an increase in quality.

• The enterprise application architect at a healthcare technology company spoke about a sales development agent that enables sales reps to more quickly generate insights about opportunities, conduct deep research on that prospect, and seamlessly capture meeting notes in Salesforce. This includes integrating Copilot for Sales with the organization’s Salesforce CRM data.

• The executive leader at an engineering professional services firm said their company built marketing and sales agents to engage with prospects and customers via chat and email to assess propensity to buy. The interviewee said: “The next phase is to identify qualified buyers and create proposals that match their needs. Other agents will validate a proposal’s compliance against the requirements and confirm that our corporate standards are met.” They also estimated win rates could increase by 10% to 15%.

• The head of applied AI and practice at a financial services organization described a bespoke agentic interface used by the entire company. They explained that salespeople use it when preparing for sales meetings because it pulls together public and in-house data to create bios for all meeting participants, a review of what products and services a customer is already using, and recommendations about what to pitch in the meeting.
  The interviewee noted that reducing the time needed to gather all of this information from hours to less than 5 minutes improved quality and comprehensiveness. They said: “While our lines of business are complementary, the typical client is only doing one thing with us when they could do a few. Our employees don’t know all our product sets. We can now sell across all our portfolios.”

• The CTO at a technology service professional services organization talked about a variety of agents to improve the sales function: “We have built RFP and proposal generation agents that facilitate RFP responses. It is currently being piloted, but all 450 sellers should benefit from it, [which would create] better responses and cut the time to generate them by up to 40%.”They also said: “Sellers who are using our agents are outselling other agents. The agents review an account — including the current technology landscape, industry data, and other insights — so a seller can start to plan the next best thing to sell a client. … The recommendation agent creates three times to four times more sales ideas than a person on their own.” The interviewee also said their company increased win rates for sellers using the agents by five percentage points: “Currently, the top 25% of our sellers are all using the agents. We now need to get everyone else to use the agents to make them better sellers.” Expanding on the impact of their organization’s sales agents, the interviewee said: “Sellers who use our agents and dashboard are selling 65% more in total sales and have 73% larger pipelines. They come up with more sales opportunities and get them qualified faster.”

• The enterprise application architect in healthcare technology said their company is building agents for the contact center, and they believe customer service employees will have better access to customer and company information to resolve issues faster and improve customer satisfaction.

• The head of colleague productivity and technology at a media and entertainment company described using agents to provide the organization’s client services teams with comprehensive views of their clients to improve interactions. They said, “We are reducing the admin overload so people can spend more time with clients.”

• The managing director of strategic supplier relationships and AI transformation at the same media and entertainment company said agentic AI “will have a big impact on customer retention,” and estimated their organization’s churn rate could decrease by one-third.

• Survey respondents reported:

  • 7.2% more opportunities generated.⁷
  • 7.3% improvement in opportunity to lead conversion rate.⁸
  • 5.0% improvement in win rates.⁹
  • 3.8% decrease in sales cycle duration.¹⁰
  • 9.0% improvement in customer retention.¹¹
  • 6.5% improvement in customer satisfaction (CSAT) scores.¹²

See Appendix D for more survey findings.

Modeling and assumptions. Based on the interviews and survey, Forrester assumes the following:

• Before using Copilot Studio, the composite’s annual baseline revenue was $6.25 million, and this is held constant for model simplicity.

• The increase in the number of qualified leads improves over time as more agents are built and their sophistication increases.

• Creating more successful agents that more salespeople use increases the composite’s win rates.

• As agents are deployed to assist customer service representatives and to provide better self-service to customers, the composite’s customer retention rate improves.

• Forrester applied a pre-agentic AI net margin of 7.33% to the increased revenues.¹³

Risks. Results may not be representative of all experiences, and the benefits will vary between organizations based on the following:

• The number and types of agents built.

• The organization’s level of IT sophistication.

• The organization’s industry and region.

Results. To account for these risks, Forrester adjusted this benefit downward by 15%, yielding a three-year, risk-adjusted total PV (discounted at 10%) of $5.7 million.

Go-To-Market Transformation: Product And Market Innovation

Evidence and data. Interviewees said using agentic AI transformed their organizations’ go-to-market capabilities through better and faster product/service innovation and the ability to serve new markets because of automation and efficiencies created by AI agents. This increased revenues and also created more meaningful work opportunities for individuals across many functions, including sales and marketing, product strategy and development, analytics, and delivery teams.

• The managing director at a media and entertainment company described how their organization is using agents to unlock the value of historical information: “We have 40 years of interviews. We can now use that history to target small and medium-sized enterprises. Agentic AI enables us to create a lighter version of what we do with our large customers. This requires a different approach that is only possible because of AI.” This individual estimated that, in a couple of years, revenues from this new offering will be worth 4% of current revenues.

• The chief AI officer at the same organization explained that agentic AI democratized data analysis and insights for more users working with customers, which improved the value delivered to those customers, increased revenue, and improved customer retention. They said: “A user may ask an orchestrator bot a question to understand what is happening in a [consumer market]. The orchestration agent will work with specialized agents to look at different cuts of the data and come up with a go-to-market plan.” The interviewee said their organization’s in-house team is initially using the tools, but that the longer-term goal is to make them available for customer self-service.

• The same interviewee also spoke about using agents to improve advertisements: “Different agents will work together on storyboarding and creating different creative elements. Other agents will then analyze the ad concept and suggest improvements. The goal is for agentic AI to get the effort to the 80% mark, and humans will do the rest of the work.”

• Another agentic-AI innovation at the same organization is creating virtual focus groups in which agents are avatars of different human personas.

• The executive leader at an engineering professional services firm spoke of many agentic-AI projects to foster innovation and accelerate delivery. This was described as the cost of quality. They said, “We will help people focus on the creative elements of their work, which will improve customer and employee satisfaction and deliver more value.”

• The same interviewee also estimated that innovation in how work is done will increase revenue per head by 20% to 30% over the next three years.

• The CTO at a technology professional services organization said: “[Agents] are helping my company sell to smaller organizations. This requires new ways of working and a lower price point. We have agents to support delivery teams and some client self-service tools. Agentic AI changes the way we think about delivery.”

• The enterprise applications architect at a healthcare technology company said, “We are building agentic AI into our products to create new services and new revenue streams.”

• Nearly half of marketing and sales survey respondents reported improved agility and faster time to market as a benefit.¹⁴ Seventy-one percent of marketing survey respondents said they believe customer insights will improve significantly.¹⁵ And sales respondents expect a 2.7% increase in average deal size.¹⁶

Modeling and assumptions. Based on the interviews and survey, Forrester assumes the following:

• The composite’s increase in revenue improves over time as more products and services are brought to market and new market segments (e.g., organizations of different sizes, regions, market-adjacent areas) are able to be served.

• Forrester applied a 7.33% pre-agentic AI net margin to the increased revenues.

Risks. Results may not be representative of all experiences, and the benefits will vary between organizations based on the following:

• The organization’s industry.

• The products or services the organization previously delivered to customers.

• The company culture in terms of embracing technology for innovation.

• Any privacy/contractual concerns around using existing datasets.

Results. To account for these risks, Forrester adjusted this benefit downward by 15%, yielding a three-year, risk-adjusted total PV (discounted at 10%) of $2.4 million.

Operations Transformation: Labor Efficiencies

Evidence and data. Interviewees said that by streamlining and automating processes with agentic AI, employees save time on tasks that are often low-value and repetitive. These efficiencies can result in two types of labor savings: saved time by existing employees that’s reallocated to higher-value work and agentic AI efficiencies and skills/knowledge augmentation that result in not needing to add or replace people to complete all the work previously done by human employees.

• The executive leader at an engineering professional services organization said they believe some of their company’s human labor will be replaced by digital workers: “Traditionally, revenue growth has been based on labor growth. What is changing is the labor mix. The direct cost of agents is another form of labor cost. Instead of growing our headcount by 10% to achieve 10% growth, we may only need to grow headcount by 5% and spend some more money on technology.”

• The same interviewee estimated that the time their organization needs to create a proposal can be reduced by 50% and that the total cost of production for an RFP response will go down by 20%. The company currently spends several hundred million dollars biding on work.

• The head of applied AI and practice at a financial services organization said a few thousand client service team members use the company’s agentic AI platform to gather information for sales meetings and create meeting content, and that the first draft can now be done in minutes rather than hours.

• Th chief AI officer at a media and entertainment company shared an example of using AI for data analytics: “In the past, someone needed to ask the data science team to do the analysis, even for something as simple as an SQL (structured query language) query. What could take three days of elapsed time can now be done in minutes.”

• The same interviewee also shared that teams gained the ability to test new product ideas faster: “If there are 50 ideas to test for a client, it can take weeks or months. Using avatars for testing, our teams will be able to complete the work in a single day.”

• The enterprise application architect at a healthcare technology company described customer support and engineering teams using an agent to do incident root cause analysis, delivering a 25% productivity gain.

• The head of applied AI and practice in financial services said their organization created an agent to do payment validation that takes action, is autonomous, and has a digital identity just like an employee would: “There are a few thousand payments per day that need to be validated, and the agent takes care of 90% of the work.”

• The same interviewee said their organization also provided agentic AI to the legal team for vendor contract review. They said: “The team is saving at least 25% of its time and has also reduced the spend on external counsel.

• The enterprise applications architect at a healthcare technology company said agentic AI saves time for their company’s legal team: “Information search has the potential to happen 25% faster based on the pilot.”

• With regards to project delivery, the executive leader at the engineering professional services firm said embedding agentic AI in software will make technical delivery easier and that their company’s goal is to buy these solutions rather than build them.

• The CTO a technology professional services organization explained that pairing agents with people is as much about quality as it is efficiency: “We are getting higher quality out of our legal teams. There isn’t any part of our processes that won’t benefit. We’ve created a quality assurance checklist agent that is more comprehensive than the one the human workers use. Faster and higher quality are both big benefits for us.”

• The same interviewee also credits agents with an expected 20% to 30% reduction in coding effort. They said, “Without agents, better training of employees might delivery only a 5% to 10% reduction.”

• Regarding the IT area, the information governance manager at an energy company explained how agents are being used in conjunction with the IT ticketing system to more quickly triage and resolve tickets: “There could be 20% to 30% time savings, which affects not only IT, but also all the employees submitting tickets.”

• Survey respondents said they expect to see the following:

  • A 6.8% reduction in marketing cost per lead.¹⁷
  • Each seller pursuing 6.6% more leads.¹⁸
  • Each customer service representative handling 6% more cases.¹⁹
  • A 4.7% reduction in customer service resolution time.²⁰
  • A 4.8% increase in first-call resolution rates.²¹
  • A 3.9% reduction in average call time.²²
  • A 16.4% reduction in cost per finance analysis request.²³
  • A 12.5% reduction in finance deal review time.²⁴
  • Each HR representative handling 5% more inquiries.²⁵
  • A 10.3% reduction in HR issue resolution time.²⁶
  • An 8.7% reduction in IT operations costs.²⁷
  • A 20.7% reduction in legal cost per review.²⁸
  • A 9.9% increase in order accuracy rates.²⁹
  • An 8.5% increase in successful delivery rates.³⁰
  • A 4.6% increase in SLA compliance rates.³¹

Modeling and assumptions. Based on the interviews and survey, Forrester assumes the following:

• Forrester conservatively estimated the percentage of positions opened up by people leaving the company that do not need to be refilled because remaining employees can complete the remaining work in conjunction with AI agents. (See Appendix C).

• During the three-year period, the composite refills all sales and IT positions.

• Customer service has the largest attrition rate, and the composite does not refill 6% of these positions in Year 3.

• For the other functional areas, between 5% and 10% of the open positions are not refilled in Year 3.

• Over the life of the study, there are 4,938 leavers. Of this total, the composite does not refill 158 positions: 69 in customer service and 89 across all other functional areas.

• Forrester estimated different fully burdened costs (including salary, benefits, and payroll taxes) for customer service workers and used an average for the costs of all other workers.

• Because these are positions opened up by people who leave the company for reasons not related to AI and the positions are not refilled, Forrester considers these to be hard cost savings that aren’t calculated as a productivity gain.

Risks. Results may not be representative of all experiences, and the benefits will vary between organizations based on the following:

• The number of people who choose to leave the company.

• The level of efficiency achieved through the application of agentic AI.

• Average salaries.

• Country-specific regulations regarding the refilling of open positions.

Results. To account for these risks, Forrester adjusted this benefit downward by 10%, yielding a three-year, risk-adjusted total PV (discounted at 10%) of $8.9 million.

Operations Transformation: External Spend Reduction

Evidence and data. Interviewees said another way in which transforming operations with agentic AI reduced costs was by decreasing the money spent externally on services and goods. They explained that by freeing up time and providing employees with skills and knowledge they would otherwise not have, more and different types of work can be completed internally rather than contracting with external suppliers. Another, more specific form of reduced external spend is related to IT vendor solutions such as other AI solutions and other types of technology for which AI agents can be used (e.g., analyzing trouble tickets).

• The chief AI officer in media and entertainment said their company often needs to pay third parties to assist with data collection, but they expect this will be reduced with the use of agents.

• The information governance manager at an energy organization explained: “The primary differentiating factor in the energy sector is efficiently getting oil out of the ground and refining it. The goal is to make oil and gas affordable, reliable, and ever cleaner. We are looking to use agentic AI to lead in this space.”

• The same interviewee said they believe using agentic AI can reduce their organization’s external spend across all areas — including COGS — by 3% to 5% in two years. As part of this, the company is looking to use agents to “build better partnerships across the industry and better integrate across the full value chain.” One especially valuable expected opportunity is reducing a refinery’s maintenance downtime, which will be accomplished with a mix of Microsoft solutions and industry-specific AI solutions provided by other vendors.

• The same interviewee said their company is also using AI to make other IT systems (e.g., helpdesk ticketing) more efficient: “[These external systems have] significant costs associated with them that can be reduced.”

• As discussed earlier, the executive leader at an engineering professional services organization reported spending several hundred million dollars per year on bidding for new projects. They said a portion of this is spent externally and that they expect to reduce this spend by 20%.

• In addition to internal time savings for legal teams, interviewees explained that these gains also lowered spend on external counsel.

• Many of the inefficiencies described in previous sections resulted in using outside service providers and/or contractors, and that this work could be in marketing, IT, data analytics, administrative support, translation, etc. But interviewees said much of that external expenditure can be replaced by agents doing the work and/or employees performing those tasks.

• Survey respondents reported:

  • A 4.7% reduction in external marketing agency spend.³²
  • An 11.9% reduction in outsourcing spend for the finance organization.³³
  • A 6% reduction in outsourcing spend for the IT organization.³⁴
  • A 5% reduction in other AI technologies spend.³⁵
  • A 6.4% reduction in external counsel spend.³⁶
  • A 10.4% reduction in procurement costs.³⁷
  • A 5.1% reduction in days of inventory held (which can incur external costs such as warehousing and financing).³⁸

Modeling and assumptions. Based on the interviews and survey, Forrester assumes the following:

• Forrester conservatively estimates the composite spends 8% of its baseline revenues on external goods and services excluding COGS.

• The size of the external-spend reduction increases over time, commensurate with the level of agentic AI adoption and the amount of employee time savings.

• Approximately 0.6% is spent externally on software and systems. This is based on Forrester research showing that, across all industries, 2.8% of revenue is spent on technology, and 22% of that is spent on software.³⁹

• The savings in this area increase over time as the composite uses agentic AI to replace or reduce other systems and reduce the need for AI solutions from other vendors.

Risks. Results may not be representative of all experiences, and the benefits will vary between organizations based on the following:

• The amount of outsourcing.

• The amount of outsourcing spent on IT.

Results. To account for these risks, Forrester adjusted this benefit downward by 10%, yielding a three-year, risk-adjusted total PV (discounted at 10%) of $16.2 million.

People And Culture Transformation: Reduced Employee Attrition

Evidence and data. The final transformation area is people and culture. Interviewees explained that, within this, the first quantified benefit has been fewer employees voluntarily leaving because they’re happier and have better goal obtainment that results in higher compensation (e.g., salespeople hitting sales quotas, other workers receiving full bonuses). People and culture transformation typically lags behind go-to-market and operations transformation because it takes time for a company’s culture to change and for technology to impact employee satisfaction.

• The information governance manager at an energy company said agentic AI can reduce attrition by improving workplace safety: “Agentic AI will have a huge impact in terms of safety [by] saving lives and reducing risk.”

• The same interviewee said their organization is building HR agents to facilitate employee self-service and make it easier to for people to do things important to them (e.g., booking and managing vacations). These agents integrate into the company’s various payroll systems around the world and other vendor solutions including IT service management and human capital management systems.

• The executive leader at an engineering professional services organization said, “I expect to see an increase in employee satisfaction from doing more meaningful work.”

• Each interviewee generally said agentic AI empowered people to focus on higher-value work, which increased employee satisfaction.

• Survey respondents said they expect a 10.9% improvement in employee retention.

Modeling and assumptions. Based on the interviews and survey, Forrester assumes the following:

• The composite’s initial employee attrition rate prior to using agentic AI is 17.4%.

• With agentic AI, the composite’s employee attrition rate decreases by 5% (0.87 percentage points) in Year 2 and 10% in Year 3.

• Based on separate salary assumptions for customer service (primarily contact center) employees and all other employees, the average cost to hire a new employee is 30% of the blended fully burdened cost of an employee.

• As the headcount mix changes because not all leavers’ roles are refilled, there is a change in the blended fully burdened cost.

• The composite doesn’t see this benefit in Year 1 because it takes time for employee satisfaction and employee retention to improve as the technology is adopted and the culture changes.

Risks. Results may not be representative of all experiences, and the benefits will vary between organizations based on the following:

• The organization’s employee attrition rates.

• Use cases that improve employee satisfaction and outcomes.

• General company culture.

Results. To account for these risks, Forrester adjusted this benefit downward by 5%, yielding a three-year, risk-adjusted total PV (discounted at 10%) of $4.9 million.

People And Culture Transformation: Reduced Onboarding Costs

Evidence and data. Interviewees said their organizations realized other quantified benefits of people and culture transformation even more quickly: reduced onboarding time and decreased costs of new hires and people changing roles. There are many agentic AI use cases for knowledge management and HR processes to facilitate onboarding along with related use cases for knowledge capture when someone leaves a company.

• The managing director of strategic supplier relationships and AI transformation at a media and entertainment company described a range of agentic AI use cases to support joiners, leavers, and transfers. For new hires, the organization provides agentic data mining tools, and the interviewee said this will “reduce the time to be fully productive from six months down to two or three months.” They also estimate the time required for transfers will be reduced from two months to one month.

• The same interviewee explained that another benefit of these knowledge management agents is supporting international operations: “As an MNC (multinational corporation), we have people around the world doing similar work. Someone in the US might benefit from questions another employee is asking in China.”

• The information governance manager at an energy company spoke about a use case targeting leavers. They said the goal is to make sure institutional knowledge is not lost when someone leaves and to then disseminate that knowledge across the entire organization.

• The CTO at a technology professional services organization said their company’s agentic work helps onboard salespeople 50% faster.

• Seventy-five percent of surveyed HR decision-makers said they expect agentic AI to improve recruiting and onboarding.⁴⁰ Fourteen percent expect a reduction in the onboarding costs for new hires.⁴¹

Modeling and assumptions. Based on the interviews and survey, Forrester assumes the following:

• The composite’s onboarding time savings increase as it creates better knowledge management and HR process agents and, by Year 3, the time to onboard new hires and transfers is reduced by 50%.

• The value created by someone in a new role — whether they’re a new hire (F5) or a transfer (F12) — is less than the value of someone who if fully proficient in that role.

• Forrester applied a 50% productivity capture to new hires (F6) and movers (F13) because not all time saved translates into additional, value-add work being completed.

Risks. Results may not be representative of all experiences, and the benefits will vary between organizations based on the following:

• The number of new hires and movers.

• The time it takes to onboard.

• The nature of the work/role.

• Average salaries.

Results. To account for these risks, Forrester adjusted this benefit downward by 5%, yielding a three-year, risk-adjusted total PV (discounted at 10%) of $6.5 million.

Unquantified Benefits

Interviewees and survey respondents mentioned the following additional benefits that their organizations experienced but were not able to quantify:

• Improved security and compliance. Interviewees generally felt that using Microsoft’s agentic AI solutions supported their organizations’ security and compliance efforts. They attributed this to integration between agentic AI development tools, first-party agents, and the broader Microsoft stack, including Azure security capabilities and Microsoft security solutions such as Purview for information protection and compliance.
  
  Another leading factor was a reduction in the use of shadow IT that creates security and information leakage risks. Fifty-eight percent of survey respondents said information security will improve significantly with Microsoft AI agents. The head of applied AI and practice at a financial services organization said: “We were very proactive putting the necessary security and governance in place. Any high-value use cases require the completion of a detailed questionnaire that’s reviewed by a governance committee made up of IT, legal, risk, and compliance resources. We carefully assess what data will be used, who will have access to it, and any ethical considerations.”

• Better information quality and management. Quality data and information — both structured and unstructured — underpins the value of agentic AI. As part of agentic AI programs, companies are investing in data hygiene beyond what they may have previously done during the first wave of genAI adoption, and agentic AI can be used as part of the initial work and ongoing management or improvement of information quality. The chief AI officer at a media and entertainment company shared that the first pillar of their organization’s AI strategy is about data. “We have a lot of proprietary datasets that we can use. Bringing together datasets with clever AI will create a lot of new opportunities for us.”

Flexibility: Incremental Value Of First-Party Agents

Evidence and data. Microsoft has recently begun building and making available first-party agents designed to be used with required integration and customization work and that extend the value of Microsoft 365. Because these agents have been generally available for less than six months, Forrester modeled the value for the composite organization but did not include it in the core financial analysis.

These agents can support organizations in three ways: 1) providing capabilities that someone would not build themselves (e.g., Facilitator meeting agent in Microsoft Teams); 2) providing capabilities that organizations may otherwise build themselves, potentially reducing time to benefit and the total cost of ownership (e.g., Researcher agent in Microsoft Teams); and 3) providing a subset of required capabilities, which can reduce the time and cost to develop a full solution (e.g., Sales Agent). In some cases, a first-party agent may provide benefits in more than one of these categories.

Interviewees frequently said their organizations would like to purchase agentic capabilities wherever possible, customize/extend with low code, and then — only when necessary — build from scratch with either low-code or full-stack development tools.

• The head of colleague productivity and technology at a media and entertainment company said: “The real benefit with Researcher is when we point it at our own data, it give us a combined view from inside and from external data. … We have many use cases, including client intelligence, business development, and market intelligence. … Probably 20% of our people can benefit from Researcher.”

• The same interviewee also described seeing early success with the Facilitator agent: “We have 130 people testing Facilitator, and they all think it is very useful. Everyone previously liked that they were able to get a meeting summary, but Facilitator tracking action follow-ups is really powerful. Nobody has to log actions, which saves time. It also helps people stick to the agenda. I think we can reduce every meeting by 10 minutes, which saves a lot of time and frees up rooms.”

• The enterprise applications architect in healthcare technology said their organization uses Research Agent for competitor analysis, sales prospecting, and assessing existing accounts: “I have heard about productivity gains of 20% when using Researcher. Personally, I love it.”

• The chief AI officer at a media and entertainment organization described their team’s use of the Analyst agent: “I can do many of the same queries using tools like Python, and this saves a lot of time. My team consists of 95 people, and about 20% are using Analyst. They are probably saving 30 to 45 minutes per day.”

• The information governance manager at an energy company said they’re very enthusiastic about the Analyst agent and its Python integration with Excel because the energy sector is very data analytics-intensive, and there are a lot of people across various analyst and finance teams who can benefit. They said: “[The Researcher agents is] strategic in nature, so it can be used by people across the value chain. We have highlighted the value of this to our senior management team.”

• The CTO at a technology professional services organization explained that much of the value of first-party agents for their company comes from how they integrate with other Microsoft solutions: “Researcher and Analyst are extremely powerful because they tie into the Microsoft stack. We are using them in conjunction with Dynamics CRM, Sales Copilot, and Sales Agent.”

• Survey respondents said that within the next two to three years, they expect 45% of the total value of agentic AI to come from agents offered by Microsoft.

Modeling and assumptions. Based on the interviews and survey, Forrester assumes the following:

• The composite sees no incremental improvement to its go-to-market transformation revenue because it can build much of what first-party agents do in-house.

• The composite improves its operational transformation because it uses first-party agents (e.g., Facilitator) to do things the organization can not build.

• It reduces external costs through the efficiency gains and by reducing spend for AI and other systems.

• The composite’s use of first-party agents reduces its spend across all three cost categories discussed later in the study.

• The incremental costs associated with integrating and managing first-party agents is equal to 10% of the total benefits achieved (G12).

• Each of these benefits ramps over time as Microsoft releases more first-party agents and enhances those already available and as the composite organization learns to use first-party agents.

Risks. Results may not be representative of all experiences, and the benefits will vary between organizations based on the following:

• Goodness-of-fit of first-party agents.

• The organization’s buy versus build preference.

• The total values of all other benefits and costs.

Results. To account for these risks, Forrester adjusted this benefit downward by 10%, yielding a three-year, risk-adjusted total PV (discounted at 10%) of $3.4 million.