LICENSED FOR DISTRIBUTION
Where You Should Use Artificial Intelligence — and Why
Published: 03 July 2017 ID: G00328113
Analyst(s):
Whit Andrews
Summary
Organizations are exploring AI's application for its own sake rather than based on a need for "novelty." This research analyzes the potential of AI in various horizontal and vertical use cases for CIOs to digest before commencing their own AI projects.
Overview
Key Findings
Interest in artificial intelligence (AI) has accelerated markedly; the volume of client conversations on the topic increased 200% from 2015 to 2016, and the most popular topics currently center around the nature of AI and why its fortunes have improved.
Organizations typically are hunting for justification to explore AI. This is often driven by top-down interest in whether the technology should be considered for its own sake, or if it is applicable in situations where a business challenge has proven intractable.
Media and intellectual celebrities tend to incorrectly characterize AI as a way of replacing people in workplaces with robots.
AI is particularly applicable in situations demanding swift or otherwise very large scale classification and prediction, especially when data is well-defined and of good quality.
Recommendations
CIOs responsible for AI-powered projects should:
Shortlist applications that can improve through training. Interpret their value using traditional product evaluation strategies such as RFIs, RFPs and proofs of concept (POCs). Select and prioritize AI projects according to the business value they will deliver — do not force-fit projects for the sake of doing AI.
Employ AI to solve challenges in which you lack the resources or corporate worker base to succeed. Identify repeatable tasks where the need is repetitive but the outcomes vary, for which AI is particularly useful (for example, frequently asked questions without frequently identical answers).
Combine your efforts with those of chief data officers (CDOs) in evaluating whether a project is good for AI, based on the data the applications will ingest and the validity of data capturing the outcomes of the project.
Analysis
Employ AI to Solve Challenges in Which You Lack the Resources or Staff to Succeed
Rather than their experimental or "cool" value, AI projects and pilots should earn their priority based on the needs of organizations considering them.
What makes the best AI projects stand out is that they allow for solutions that previously would have been impossible to conceive, because they include what seems like human insights but at a volume humans could never achieve.
In some cases, high-tedium and medium-value work, such as searching for all documents about a client organization every time an interaction takes place, will occur more reliably with the automation process's AI-driven improvement. This would be more likely to surface the most relevant documents to the interaction of the moment.
We recommend to organizations in any industry that AI be employed to do what they already need doing, more than for gee-whiz "moonshots." In different industries and departments, AI's particular value can be applied without treating it as a way to solve every problem.
In general, AI improves the practice of connecting executives and workers to computing resources with previously impossible simplicity. The permutations of virtuosic computer-driven classification and prediction — the essence of AI, and what makes it "seem human" to users — result in a wide variety of use cases that in many ways seem just to recapitulate those of computing in general.
Big data was famously defined through the volume, variety and velocity of the data that organizations found themselves facing. AI can be seen as converting such overwhelming information and data flows into granular insights on which AI-driven systems may take automated actions. Data about customer behaviors and interests might have been overwhelming until advanced analytics strategies could parse them and make them useful, but AI now offers the chance not just to analyze the aggregate, but respond in the particular. Crucially, AI does so in a manner that allows it to contribute to improvements in its own performance.
Case Studies
We recommend that organizations that explore applications using virtual assistants consider the potential available when customers (or constituents or workers) employ such automated, lower-friction methods of interacting.
Deakin University employed an automated AI-driven immediate question-answering system that led students to ask questions they would not have posed before, such as where to get something to eat immediately. The system now offered students an impersonal, always-on and always-present advisor. This new volume of requests was not disruptive because of the capacity of the project's interactive assistants to respond, and it led to new behaviors and satisfactions in the student body.
Telefónica used AI to route requests for help from workers to other workers. Being an organization where acquisitions had significantly increased process complexity, this allowed it to connect people who would have had no way to find each other previously.
Most organizations are still only beginning their explorations of projects that will particularly apply AI. Gartner recently canvassed about 80 members of its Research Circle to get a sense of where they stand on AI projects. Most in this sample said they are still gathering knowledge and developing their strategies for applying AI (see Figure 1).
Figure 1. Current Stages of AI Adoption Strategy
Source: Gartner (July 2017)
Meanwhile, AI continues to seep into other applications as the trained capabilities in machine learning multiply in application categories devoted to advanced analytics, customer experience and the digital workplace.
Such self-teaching (or self-learning) is not purely automated either in selecting strategies to employ (such as which algorithms to add to an analytical cocktail) or even in calculation tweaking. While neural net programming is delivering automated improvements in academic and controlled environments (such as game-playing or training datasets), organizations must expect that they will continue to engage with AI for the foreseeable future. They must supervise how the systems improve their outcomes and whether the tactics they propose are best. Systems that use machine learning sometimes must be retrained, and common sense is still necessary to select improvements. Tellingly, DeepMind's AlphaGo — which has been retired from competitive Go play — is increasingly being used by players to invigorate play styles in the international community, not merely as a demonstration of computational force.
Select Business Challenges Based on Their Suitability for AI
Vendors sometimes pitch new technologies first only as solutions for new problems; time generally reveals them initially to be most useful for addressing older challenges in innovative ways. Thus did "e-business" turn out to not radically reinvent many companies' businesses initially, but provide them first with a means of reducing errors in orders. What followed was a means of increasing the dynamism in business relationships, and then ultimately, perhaps, a means to reinvent how they did business.
AI is similar, in that it embodies previously unimaginable uses for technology and presents them to executives who can employ it for ordinary challenges.
As a result, the projects that make the most sense for organizations — regardless of the specifics of the verticals involved as well as the applications involved — include the following:
Data whose quality and credibility are good, and which have sufficient scope to fully address the problem. The more layers of abstraction that are introduced between data and analytical outcome, the harder it is for common sense to reveal error in a timely fashion. Self-improvement is not fully automated, but it can be swift and difficult to parse. We recommend organizations invest in efforts to verify and improve the quality of data for such purposes with particular strength (see "Preparing and Architecting for Machine Learning" ).
Data with clear parameters that are reliably followed. Any analytical strategy relies on the ability to establish a multivariate view of the data it is analyzing, and the ability to reduce elements such as transactions or interactions to parameterized vectors is fundamental. Too few such parameters, or blurriness or inconsistency in their content, are insurmountable obstacles.
Demonstrable value to business stakeholders based on historical desire.Don't aim for the moon if what you need is a satellite in orbit. Initial pilot projects should target existing needs, and initial projects should follow in the pilot projects' paths. Review existing projects or those just beginning flight for AI applicability, and make AI part of the consideration for future projects for primary function or for extra benefit.
A foundation demonstrating that the organization's goals are reasonable and possible. In the best case, organizations will discover similar projects completed previously by participants. This is especially important when using service providers, which should be able to provide examples of very similar or somewhat similar past projects. Lacking such proof, organizations should seek out academic research indicating that the goals are reasonable.
Shortlist Applications That Can Improve Through Training
"Does artificial intelligence really work?"Organizations who look to AI to solve a challenge they face often begin their quest with the need to understand whether AI has "crossed the Rubicon" into functionality — and understandably so. Famously, AI is described as a technology that is not achievable — whenever it meets the conditions that were agreed upon for its fulfillment, academic researchers impose new conditions.
To address this, we recommend that organizations approach their AI projects with the same deliberation and care they place on conventional projects that have no AI associated with them. The key differences for which they should prepare are the issue of improvement through training and the lack of immediate visibility into how algorithms in AI perform. In other words, AI-powered applications that work well may improve over time and through investment in training, but understanding why may not always be possible (see "Innovation Insight for Deep Learning" ).
Organizations should follow ordinary procedures in seeking AI-powered applications or evaluating adding AI capabilities to existing applications. They should be sensitive to the fact that they will face the same project concerns in any AI project, including integration to existing Mode 1 and Mode 2 applications, which will serve as sources of data and vehicles for output from the AI elements.
In evaluating such solutions, organizations should use sets of conventional requirements that each vendor should be measured against. AI solutions may (and should) improve over time, but organizations should employ real metrics — as opposed to potential or hypothetical metrics — to assess and evaluate them.
Such procedures do not mean that projects will themselves be routine. On the contrary, AI projects are still experimental, and should be seen as an investment in developing understanding of the technologies. In particular, projects that employ unsupervised learning or other means of machine learning intended to drive outcomes should be considered experimental. The greater the supervision, the more ordinary project governance can be. Project practices should reflect research goals.
Defer to CDOs in Evaluating Whether a Project Is Good for AI
The CDO is accountable for managing information as a strategic asset. From this position, his or her value is central to every aspect of the AI project. The CDO is responsible for the organization's data architecture, its quality and its governance, as well as for data's provenance and credibility (see "Chief Data Officer Desk Reference for Artificial Intelligence" ). The CDO is as important as the CIO in determining how to employ AI in organizations, because he or she is closest to the data that will feed the AI-powered applications. The CDO is most likely to have solid connections and authority in the views of the business analysts and data scientists who are key to AI projects.
CDOs are qualified to evaluate analytical projects of all kinds to evaluate proposed methods of solving them. They should take responsibility for evaluating what projects are appropriate to address through AI, and determining how the data that will feed or be gathered as part of that project may be relied on (or not) to most effectively interpret it. Many organizations do not take this attitude.
In a recent Gartner Research Circle survey on AI plans, only two of over 80 respondents said that the CDO had initiated their AI efforts or was responsible for technology decisions, whereas about one in three respondents indicated that the CIO was responsible for those things. Those latter organizations should ensure that their top data and analytics leader is engaged.
Purposes to Which Organizations Should Apply AI
AI's Value in Application Categories
AI is a force multiplier for workers who ease the relationships between customers and organizations. AI is the engine behind rich virtual assistants that organizations seek to deploy as an initial phase of interaction between customers or constituents and the information or action that they need. Such assistants improve over time with retraining and outcome measurements. Advanced analytics have powered improving knowledge management (KM) and search for workers who face customers in sales and service. Where AI promises to shift this experience is in taking action, presenting an organizational profile that can initiate tasks on behalf of customers and prospects.
Of course, AI does not enter such realms with no existing applications. Organizations in Gartner's Research Circle indicated that they have or will integrate AI to their existing systems in customer engagement practices. The three most often-cited application categories for such integrations are all related to customer interactions:
One in three of organizations we surveyed said they will link AI to customer engagement applications
Three in 10 said they will integrate AI to call center service and support
One in four said they will integrate it to digital marketing
Indeed, Gartner recommends such applications for AI because of the very large quantity of data in the task set for organizations, because of its heterogeneity, and because of the potential for return on investment. See Figure 2 for these results and others.
Figure 2. Application Categories for AI Integration
Source: Gartner (July 2017)
Specifically, in particular aspects of customer relations, organizations should consider:
SALES AND MARKETING
Developing strategies for salespeople to execute in interactions with prospects, and matching prospects to such sellers, is a common application for advanced analytics technologies, and an appropriate way to apply AI. AI's ability to incrementally develop improvements when retrained and guided, based on a broad spectrum of aspects of any prospect or selling situation, makes it particularly apt for such applications.
Sales interactions are founded on strategies and means that are particular to given salespeople as well. Matchmaking between them and prospects or customers can increase positive outcomes without relying on cumbersome rules about the nature of the rep or the demographics of the customer.
Ultimately, automating sales processes also will be a target opportunity for AI, both through the use of virtual personal assistants and through conventional process automation. Organizations will reduce the time a human worker spends in the sales cycle. Such shifts will have significant ramifications from regulatory and legal perspectives, especially in financial services.
Recommendation: Invest initially in the use of AI and machine learning in automating the personalization of marketing and merchandising messages.
SERVICE
Initial implementations of virtual customer assistants should focus on a small percentage of customer interactions that are easy to automate. Indeed, such installations need not even be "artificial intelligence," but can rely on human-developed rules that are curated to allow questions either to be routed to a frequently-asked-questions system ("How do I turn the widget on?" ) or shifted to human representatives.
Subsequent improvements can leverage the interaction records from existing systems to train AI analyses of those records. Reverse engineer the best ways to address customer needs in an automated fashion, through textual analysis of the interactions that happen now between customers and customer service representatives. Develop question-answer pairs and complexes through evaluation of success.
Ultimately, AI will be useful in service as a way of supporting existing customer service organizations through routine interactions that are easy to define, set parameters around and duplicate.
Recommendation: Choose use cases with the greatest possibility of automating away the lowest-value transactions. Begin with rule-driven interactions and develop machine-learning-driven improvements with volume and familiarity (see "Seven Decision Points for Success With Virtual Customer Assistants" and"Four Uses for Chatbots in the Enterprise Now").
OFFICE AND COLLABORATION SUITES
AI is likely to remain largely invisible in organizations' workplaces for the immediate future, but its impact will ultimately be significant all the same. Most office workers will experience these impacts through the suites of applications that allow them to create, manage and edit the unstructured content at the heart of knowledge work.
Through 2022, AI will have minimal impact on most office employees, and their primary exposure to AI will come through cloud office suppliers such as Microsoft and Google.
However, we expect to see some conveniences sooner. Virtual assistants for resources such as office supplies or office resources (such as meeting space) will multiply and become commonplace as they are added to the productivity suites and to outsourcing arrangement applications, such as digital commerce interfaces.
Additionally, we expect to see largely mute support elements added to meetings to automate processes such as note-taking and minutes-editing, as well as to plan and prepare for future meetings.
AI will also affect other aspects of organizations through surfacing insights to people holding operational roles from data it gathers in workplaces. Examples of such operations include, inevitably, compliance warnings to workers and their managers, as well as executives outside their departments. But it will also include evaluation of retention risks and the development of automation paths and systems intended to mitigate tedium. Further, AI will improve results in KM and expertise locations — indeed, automated relevancy-improvement strategies have been at the heart of search and KM technologies of many kinds for more than 20 years, although today's AI promises even more significant improvements.
A Gartner client that is a very large telecommunications company used AI-powered expertise location in the wake of a major merger to improve interactions between the previously separate workforces. This occurred at a scale that would have been impossible to make entirely human- or conventional-software-mediated (see"Maximize the Effectiveness of Office 365 and G Suite With Everyday AI" ).
Recommendation: Target cloud office suites to expose a broad employee base to AI benefits, such as information awareness, task automation and expertise augmentation. Make the digital workplace program a home for managing and educating the workforce about AI in worker-facing usage.
SUPPLY CHAIN MANAGEMENT AND MANUFACTURING
Supply chain management's plague and opportunity is very large amounts of data from a wide variety of different resources. AI's immediate promise is to rationalize the perspectives organizations can gain from improving their management and understanding of such information in what is likely to be mostly incremental improvements.
Discovering incorrect and anomalous data that errantly confuses operations is an AI opportunity; potentially correcting such mistakes in an automated fashion could add greater value still. Additionally, AI offers the ability to sieve supplier RFPs and relationships to make predictions about the probability of future performance. For example, one Gartner client reported that its largest customer intends to install listeners within its system to allow for a more effective predictive understanding of how vulnerable that customer's supply chain will be to delays in parts availability. The client should press for transparency at least into the data the listener gathers from the client's system — but also, if possible, the conclusions it proposes to the organization that mandated it.
AI could be expected to add intelligence to digital twin models, in which the digital twin can be kept more current between opportunities to level the physical and the digital through better refinements before measurement.
AI can augment supply chain planning talent by working with the planners. It can gather data on how they have made decisions and resolved exceptions, and use them as a starting point for automated decision making or generating actionable recommendations. For example, AI can help develop better supply-to-demand matching, based on training using the planner's previous approaches and then building on top of that by learning from new data and the efficacy of its own approaches.
Manufacturers report that they look to AI for numerous purposes, most involving significant amounts of data that it is difficult to analyze by hand. Examples include process improvement, in particular, and especially predictive maintenance and similarly data-rich, expensive processes.
As is the case for many organizations, the most important factor for supply chain managers in employing AI will be the necessity to evaluate and certify the accuracy of the data that the AI-powered applications will rely on. The process of increasing automation also increases the necessity of accurate data.
Recommendation: Use AI initially to discover gaps and anomalies in communications and processes with suppliers, and work toward correction of such problems (see "Artificial Intelligence in Supply Chain: Pursue Short-Term Benefits and Mitigate Future Roadblocks" ).
AI's Value in Vertical Markets
COMMUNICATIONS
Because they are so large and experience such a tremendous load of customer interaction volume, communications service providers (CSPs) benefit largely from the conventional applications of AI to which most verticals initially are turning. Examples include contact center interactions, retail use, and fraud detection and management.
However, CSPs also have particular use cases. For example, network planning and engineering as well as network management are uniquely large-scale challenges for such organizations. To the extent that they can be simulated digitally and therefore subjected to various machine-learning disciplines with an attempt to optimize them, AI presents significant promise (see "How CSPs Can Exploit Artificial Intelligence" ).
Recommendation: Identify narrow, routine customer interactions. Test the ability of virtual customer assistants to address them; ensure that the first capabilities are stable before moving on. Gartner predicts that poor-quality implementations could significantly increase customer dissatisfaction.
GOVERNMENT
Government organizations are expressing significant interest in the use of conversational applications such as virtual assistants. They hope to ease the continued pressure for constituent service and improved interactions between the public and governmental entities — as well as inside such entities, where workers can benefit from systems that provide insights into processes and regulations.
AI systems require significant amounts of well-defined, clear data with scored outcomes attached. At least initially, for many organizations such data will be easier to find in internal than external applications, because of the comparatively robust parameters around the data. More-futuristic applications will involve integration to consumer virtual personal assistants. In Utah and Mississippi, for example, practice exams for driving tests may be completed on Amazon Echo devices, increasing the efficiency of official in-office testing through improving the preparation of applicants (see "Is Your Digital Government Platform Ready for Virtual Assistants and Chatbots?" ).
Smart city applications require more-sophisticated, multientity cooperation (or at least interoperation), and will be out of reach for many governments for the immediate future. However, the growing presence of IoT endpoints will present government organizations with increasingly large and complex data flows, which will be a challenge to leverage. This is particularly true in areas such as transport, policy/public safety and disaster management/recovery. Video analytics will grow as an aspect of AI applications in government swiftly, as performance improves and the applications that seek to employ such analytics results multiply.
Virtual assistants may offset simple inquiry loads where governments are dealing with high volumes of routine inquiries — although nonroutine inquiries should still be shunted to workers who can help with harder circumstances. Questions like "When does the office close?" will be easy for AI to handle, and even to add, "…but the last applicant must arrive 60 minutes prior to that time." AI-powered chatbots will be less likely to answer more-particular questions, such as, "How many fish can a child catch in a derby?"
Recommendation: Organizations initially should invest in smart advisors for internal applications and for intergovernmental applications, with the potential to expand such applications to include constituents — especially expert constituents familiar with government systems.
EDUCATION
For teachers, AI will continue to diminish the value of access to knowledge and information, just as other methods of information storage and retrieval have done since time immemorial. Additionally, however, AI promises to lift some of the most tedious chores from teachers, such as evaluating student understanding of materials through grading and scoring of structured and unstructured test results.
Initial applications of AI in education will follow the most effective general-purpose uses of AI, such as virtual assistants for operational advice (how to find a classroom, how to prepare materials for a test). However, the industry is eager for the implications it has in teaching. Digital assessment promises to make evaluation faster, making feedback and improvement swifter. Adaptive learning promises to allow systems to respond to student learning styles personally but in aggregate, catching signals of how a student tends to learn in order to provide him or her with instruction styles that are particularly resonant.
Ultimately, the research aspect of academic institutions will see substantial impact from AI applications in prioritization and selection of projects. Preproject evaluation of research merit will improve with AI evaluation of the methods and goals identified. We have already seen significant investment in AI in life sciences, with the goal of improving, for example, the success rate of what clinical tests are applied.
Recommendation: Organizations initially should invest in operational improvements such as student-teacher pairing and improved evaluations of structured work.
RETAIL
Marketing and merchandising are key applications that are unique to retail entities, where the applications of AI can take advantage of very significant data resources to deliver potentially measurable value. Retailers that have embraced big data analytics may have become deft at discovering previously unknown insights. AI offers the potential to react to such insights in a more agile fashion, dividing and subdividing marketing messages to target narrower and narrower psychographics, and potentially achieving more-successful personalization in store and digital commerce.
Additionally, of course, retail benefits significantly from the potential of more-personalized and automated customer service using virtual customer assistants.
Retail is demonstrating greater familiarity and understanding of AI than some other verticals are. This is partially because it encountered similar technologies such as personalization and digital recommendations earlier than any other (with the possible exception of media, thanks to advertising) during the late 1990s e-commerce acceleration. Gartner tracked social media conversations relating to AI from the first quarter of 2015, and found that the number of such conversations grew at 20% per quarter in retail-led verticals. Growth rates in other verticals have since accelerated as AI hype has increased (see the Evidence section).
Recommendation: Organizations initially should focus on promotion selection and prioritization.
BANKING
Applications for AI in banking are diverse and growing. In banking, social media conversations about AI have found the most fertile vertical of all, and the chatter in major public social platforms is noisiest here. The volume of conversation has grown, on average, about 50% per quarter since the beginning of 2015. That's faster than — and in fact doubles — the growth rate in conversation in any other vertical. Such conversations are globally prominent, but are primarily in North America, Europe, India and Australia. (The conversations are prominent in China, Japan and South Korea as well, but are secondary to other topics.)
Most visibly, banks are evaluating and deploying AI in automated interactions between customers and the banks' systems, especially using virtual customer assistants to offload routine conversations and inquiries to systems instead of relying on human interaction time. Such "bots" might be voice- or text-driven. They also might reside on an independent platform or within a social media system, such as Facebook or WeChat.
Additionally, they are using AI in developing portfolio management for customers, analyzing the decisions and interests of investors to provide them with particular insight into their investment decisions. These advisors for portfolio mix adjustment support existing customer representatives.
Other applications in financial services are designed to aid organizations in developing operational insights into extremely large datasets at previously impossible speeds. The advanced and automated analysis that AI provides is being applied in fraud prevention in particular.
Recommendation: Create new sources of revenue and pricing models by leveraging AI to uncover patterns in client and firm data. Adapt metrics to assess the value of AI to your firm by focusing on the additional benefits AI provides, such as customer experience, productivity and flexibility.
HEALTHCARE PROVIDERS
No vertical presents greater visibility for AI than healthcare. The most vivid future use cases that organizations imagine for the technology's applicability tend to center around extraordinary advances previously unimaginable, such as automated diagnosis improvements or streamlining research processes for pharmaceuticals.
But healthcare providers, with few exceptions, will find themselves hard-pressed to gather the resources and the will to pursue such opportunities, especially at a time of unique turmoil in the market around business outcomes and political confusion. Initial applications with universal value that healthcare provider organizations should pursue will focus on operational improvements, which will improve financial performance, employee relations, patient satisfaction and quality of care.
AI excels at using very large amounts of data to establish insights into what results in anomalies and errors, making it attractive for addressing issues such as coding errors, appointment no-shows and similar costly challenges to providers. "Moonshot" ideas such as better cancer diagnosis and other opportunities do not lack merit; but for many providers, the least-visible challenges such as viable scheduling practices and other improved practices are the best investments.
Recommendation: Healthcare providers should initially focus their efforts on automating the practice of identifying errors in labor-intensive processes — especially those where mistakes are costly, to reduce risks but in particular to improve operations and costs.
Evidence
Methodology for analysis of social media conversations: We used automated social media listening tools to track users' responses on social media and public discussion forums. The period for the analysis of overall mention count was from 1 January 2015 through 31 March 2017.
"Social media mentions" denote the inclusion of a monitored keyword in a textual post on a social media platform. A high count of mentions should not be considered an indication of positive sentiment by default. Social media sources considered for this analysis included Twitter, Facebook (publicly available information only), images (comments only), aggregator websites, blogs, news, mainstream media, forums and videos (comments only). All regions and major world languages were covered for the study.
Note: The social media analytics team members who contributed to this research include Anjali Grover, Ayush Saxena and Arnav Saxena.
Online survey for Gartner Research Circle members: An online survey was held from 5 April to 21 April 2017 among Gartner Research Circle members — a Gartner-managed panel comprising IT and business leaders. In total, 83 members completed the survey.
The survey was developed collaboratively by a team of Gartner analysts, and was reviewed, tested and administered by Gartner's primary research team.
© 2017 Gartner, Inc. and/or its affiliates. All rights reserved. Gartner is a registered trademark of Gartner, Inc. or its affiliates. This publication may not be reproduced or distributed in any form without Gartner's prior written permission. If you are authorized to access this publication, your use of it is subject to the Usage Guidelines for Gartner Services posted on gartner.com. The information contained in this publication has been obtained from sources believed to be reliable. Gartner disclaims all warranties as to the accuracy, completeness or adequacy of such information and shall have no liability for errors, omissions or inadequacies in such information. This publication consists of the opinions of Gartner's research organization and should not be construed as statements of fact. The opinions expressed herein are subject to change without notice. Gartner provides information technology research and advisory services to a wide range of technology consumers, manufacturers and sellers, and may have client relationships with, and derive revenues from, companies discussed herein. Although Gartner research may include a discussion of related legal issues, Gartner does not provide legal advice or services and its research should not be construed or used as such. Gartner is a public company, and its shareholders may include firms and funds that have financial interests in entities covered in Gartner research. Gartner's Board of Directors may include senior managers of these firms or funds. Gartner research is produced independently by its research organization without input or influence from these firms, funds or their managers. For further information on the independence and integrity of Gartner research, see "Guiding Principles on Independence and Objectivity."
No comments:
Post a Comment