Predicts 2018: Procurement and
Sourcing Technology Will Benefit from
Advances in AI, Data Visualization and
Published: 08 November 2017 ID: G00343167
Analyst(s): Patrick M Connaughton, Magnus Bergfors, Desere
Edwards, Kaitlynn N. Sommers
The procurement and sourcing technology market is evolving rapidly. IT
application leaders modernizing procurement technology need to plan for the
impact of virtual assistants, cloud-based business intelligence tools and B2B
Frustration with legacy software coupled with advancements in artificial
intelligence and business intelligence will lead CPOs to revisit their spend
Amazon Business and Alibaba will emerge as the dominant B2B
marketplaces for CPOs looking to consolidate tail spend.
Virtual assistants and chatbots will become standard features in P2P
software going forward.
IT application leaders modernizing procurement technology should:
Investigate which spend analysis tool your company's sourcing organization
uses today. Has it kept pace with market advancements in data visualization,
AI, natural-language processing for spend classification and pattern
recognition? If not, IT application leaders should begin exploring other
Determine how much of your company's money is already being spent in
B2B marketplaces like Amazon Business and Alibaba. Consolidate the
spending under an enterprise account, and work with procurement to
establish a basic set of guidelines on when each marketplace should be
used. Weigh carefully the trade-offs between price, lost rebates, usability
Get educated on the short-term uses of chatbots and the mid- to longer-term
applications of virtual assistants. IT application leaders can then help
educate CPOs on the possibilities of the technology. Run a pilot with the
CPO's endorsement to showcase how procurement is rising to the challenge
of creating a digital business.
Strategic Planning Assumptions
By 2022, 50% of all legacy spend analysis software will be retired; replaced by
artificial intelligence (AI)-powered, cloud - based solutions.
By 2022, 75% of all B2B tail spend goods will be purchased in an online
marketplace like Amazon Business or Alibaba's 1688.com.
By 2022, all major procure-to-pay software vendors will embed virtual assistants
and chatbots for guided buying and self-service requests.
What You Need to Know
Procurement and sourcing applications support enterprise processes relating to
the purchase of direct and indirect goods and services. Common processes
include bid solicitation, supplier management, contracting, requisitioning,
receiving, inbound invoice processing and payment.
Overall, the procurement and sourcing technology market is at an adolescent
stage of development. Procurement teams have relentlessly pursued
technological advances to professionalize and energize their work. There have
been many successes, and some types of procurement and sourcing solutions
are now in the early mainstream.
The procurement and sourcing technology landscape has a long way to go to
reach maturity. However, some areas, like spend analysis, will not have a chance
to reach maturity before advancements in areas like AI will render traditional tools
obsolete. This will set back the maturity levels initially but will lay the foundation
for much faster growth.
The strategic planning assumptions in this report highlight the most disruptive
changes to procurement and sourcing processes and enabling technologies.
Strategic Planning Assumptions
Strategic Planning Assumption: By 2022, 50% of all legacy spend analysis
software will be retired; replaced by artificial intelligence (AI)-powered, cloud-
Analysis by: Patrick Connaughton
For many strategic sourcing teams, legacy spend analysis software has failed to
deliver on its promise. By legacy, we are referring to spend analysis software that:
1. Is often installed on-premises
2. Requires defining a rigid spend taxonomy
3. Uses a combination of auto-classification rules and extensive manual effort
to "enrich" the data before category managers have a basic starting point for
The current process is time-consuming and error-prone, and prevents sourcing
from being able to quickly respond to changing market conditions and make faster
Legacy solutions are also characterized by what Gartner calls "narrow AI."
Narrow AI consists of highly scoped machine-learning solutions that target a
specific task, such as classifying spend (see Figure 1). The algorithms chosen
are optimized for that task. These solutions are limited. They are designed to
answer a predefined set of questions, but do not unearth insights the end user
didn't know to ask about. The next generation of tools will build on this foundation
with general AI. This refers to the use of machine learning to handle a broad
range of use cases. These systems will be able to successfully perform any
intellectual task that a human could perform and would learn dynamically, much
as humans do (see "Top 10 Strategic Technology Trends for 2018" ).
Figure 1. Narrow AI's Place in the Long History of AI
Source: Gartner (November 2017)
Compounding the shortcomings of these tools, some legacy spend analysis
vendors have opted to build the data visualization layer themselves. And these
proprietary visualization and reporting engines embedded in the spend software
have fallen vastly behind what's available in today's cloud-based business
intelligence (BI) solutions. For example, modern BI tools have embraced the
concepts of creating interactive storyboards and smart recommendations that
guide users (i.e., "You did x, now try y.") and smart data discovery that will
automatically generate the most-meaningful charts. Modern BI and analytics
platforms also use a self-contained in-memory engine with minimal to no upfront
modeling requirements. This architecture allows a wider range of business users
to perform interactive analysis, without the need for advanced technical or data
Meanwhile, AI is coming of age. Gartner predicts that by 2020, natural-language
generation and AI will be standard features of 90% of modern BI platforms
(see "Critical Capabilities for Business Intelligence and Analytics Platforms" ).
Natural-language processing combined with machine learning will dynamically
identify the most relevant insights and context in data (trends, relationships,
correlation patterns). It can then automatically generate a personalized narrative
for each user in their context, to explain meaning or highlight key findings in data.
The implications for spend analysis are significant. Rather than requiring
predefined spend classification algorithms, taxonomies and manual enrichment,
natural-language processing can automatically find, visualize and narrate
important findings. Put another way, the solution will be able to not only classify
the data but also interpret it to draw out trends and recommend actions. This is
essentially the first step toward a virtual category manager.
Broader procurement analytics strategies are also forcing the discussion forward.
Many procurement organizations have already layered a cloud BI tool on top of
their sourcing and procurement applications. The purpose of this tool is to not
only do spend analysis but also to enable reporting on savings and operational
metrics like cycle times.
CPOs' eagerness for a new approach, advancements in BI visualization, and the
increasing maturity of general versus narrow AI will have the following market
New entrants into the spend analysis software space will quickly leapfrog
traditional providers by embedding off-the-shelf, cloud-based BI tools with
built-in AI capabilities. These providers will set their sights on providing
holistic procurement analytics solutions spanning spend analytics,
embedded operational metrics (e.g., savings, cycle times), predictive
forecasts and aggregated market intelligence — a one-stop shop for
category managers' analytics needs.
Some traditional providers will reinvent their solutions, and some already
have, using AI to quickly analyze unstructured data, enabling real-time
category decisions. These solutions will also bolster their visualization
capabilities, partnering with cloud-based BI providers to keep pace with
innovations. Those vendors with broader source-to-pay suites will create
holistic analytics solutions spanning the entire sourcing and procurement
process. This will include context-sensitive, real-time embedded analytics.
The other vendors (our prediction: greater than 50%) selling traditional
solutions will fail to respond to the changing market needs and become
The business is not afraid to experiment with technology, and circumvents
IT on a regular basis. Some procurement organizations even have their own
dedicated analytics team. These innovative CPOs and first movers in
procurement won't wait for the vendor's to catch up. A do-it-yourself
approach will emerge, especially at those companies that decided to throw
in the towel on using legacy spend analytics products. These companies will
look to leverage pre-existing licenses for BI solutions like Qlik and Tableau
for visualization, combined with off-the-shelf AI specialists like Narrative
Science and Yseop.
Determine which tools for spend analysis your company's sourcing
organization uses today. Are they keeping pace with market advancement
in BI visualization and AI natural-language processing for spend
Leverage Gartner research to better understand which features are most
important to compare like embedded advanced analytics, smart data
discovery and interactive visual exploration (see"Toolkit: BI and Analytics
Platform RFP" ). If the current solution is not competitive, begin exploring
other options with the business to understand its readiness for a new
solution. Otherwise, sourcing leaders are likely to procure or try and build
these tools themselves without IT's guidance.
Look to see who is already using a cloud-based BI tool like Qlik or Tableau
today in your own procurement department. These solutions have grown
extremely popular over the last few years and are often procured directly by
the business. It's likely that someone is already paying a subscription fee
that can be leveraged as a starting point to pilot these initiatives. The key
here is for IT application leaders to help provide basic building blocks to the
business, then let them run with it and experiment. There will be integration
and performance challenges to consider when using cloud-based BI tools
but do not look to solve those challenges immediately. Consider this an
opportunity to innovate and pilot accordingly.
Compare the visualization and AI capabilities between dedicated spend
analysis tools as well as general-purpose BI solutions. Weigh carefully the
benefits of pre-existing category taxonomies in dedicated spend analysis
tools versus higher levels of visualization and augmented analytics (i.e., AI-
driven insights) in analytics and BI solutions. Consider a broader
procurement analytics strategy that uses a single tool for all reporting
spanning from sourcing through to payments.
"Magic Quadrant for Business Intelligence and Analytics Platforms"
"Predicts 2017: Analytics Strategy and Technology"
"Critical Capabilities for Business Intelligence and Analytics Platforms"
"Top 10 Strategic Technology Trends for 2018"
"Toolkit: BI and Analytics Platform RFP"
"Augmented Analytics Is the Future of Data and Analytics"
Strategic Planning Assumption: By 2022, 75% of all B2B tail spend goods will
be purchased in an online marketplace like Amazon Business or Alibaba's
Analysis by: Patrick Connaughton
Chief procurement officers (CPOs) are constantly looking for new ways to
increase compliance and unearth savings opportunities through better spend
visibility. For the most advanced procurement organizations, P2P software
deployments and comprehensive, streamlined policies have resulted in a high
level of compliance and visibility. These procurement teams started by increasing
visibility to the categories with greatest spend, highest potential risk and most
strategic suppliers. Now that work is done, what remains for indirect goods is the
lower dollar, infrequent or one-off, nonstrategic spend (i.e., tail spend).
Previously, the work required to gain better visibility into tail spend did not justify
the return. Considering the infrequent nature of the purchased good, savings
opportunities through supplier consolidation were sporadic. As a result, tail spend
was mostly ignored by procurement transformation teams as they prioritized
which areas to work on.
Formerly known as AmazonSupply, Amazon Business was relaunched in 2015
and provides a marketplace for businesses to sell to other businesses. Amazon
Business offers a range of categories including maintenance, repair and
operations, IT, and office supplies. Sellers can sign up and sell to other
businesses as in the consumer marketplace, but with additional functionality that
is common for B2B transactions. These include bulk pricing, bulk/pallet delivery,
consolidated invoicing, tax-exempt purchasing and integration with purchasing
systems through the PunchOut (cXML) standard.
In its first year, Amazon Business achieved $1 billion in revenue, growing at a
20% month-over-month rate. In late 2016, it expanded operations to include
Germany (see "Vendor Rating: Amazon").
In Asia, a similar story is unfolding with Alibaba. Revenue for 1688.com, Alibaba's
B2B marketplace for buyers and sellers based in China, increased 35% in fiscal
Whereas Amazon has marketplaces in North America, U.K., Germany,
Japan and India, Alibaba has experienced growth in other regions such as Brazil,
Russia and Asia. Alibaba includes features like the ability to request quotes from
multiple suppliers and verify the identities of suppliers. 1688.com allows provision
for cross-border, SMB-level procurement with minimum purchases starting from
two items and up.
Despite rapid adoption, an open marketplace concept goes against standard
procurement practices to negotiate and enforce strategic supply agreements for
higher-value spend categories. There are other limitations as well — inadequate
support for prenegotiated contract pricing, lost rebates that other sources of
supply (e.g., p-cards) may offer, and no single global marketplace..
For these marketplaces to truly grow to meet our predictions, they have to fulfill
more than just tail spend requirements. For example, some marketplaces only
flag a category as outside of purchasing policy but does not block the purchase
altogether. What's needed is a solution that restricts open marketplace shopping
enough to comply with procurement requirements and improve the value
proposition for mission-critical spend categories.
It is important to note that an intuitive user experience is one of the most highly
weighted criteria when choosing a P2P solution. This is what opens the door for
marketplaces like Amazon Business, the inventor of the "Amazon-like
experience." While it may not meet all of procurement's requirements today, it is
only a matter of time before Amazon goes on the offensive. In doing so it will
close the gaps and provide a viable solution that does not require integration with
existing P2P software to ensure compliance. Many businesses will go directly to
Amazon Business to get started quickly. Hence they will avoid the cost of
maintaining their own catalogs and expand the percentage of spend rapidly once
the workflow is enabled.
We predict that 75% of all indirect goods purchases will be transacted on one of
these platforms by 2022, due to:
Rapid growth over the last few years
Ease of access
Intuitive nature of purchasing
Continued development of features to meet procurement's specific spend
policy/control requirements. The vendor's investments in vertical-specific
marketplaces and deeper interoperability with P2P software vendors will add
fuel to the rapid acceleration
Determine how much money is already being spent with Amazon Business
and Alibaba. If the amount is significant (i.e., greater than 5% of total spend),
look to consolidate the spend under an enterprise account. Start with the
most common spend categories like office supplies, IT and maintenance,
repair and operations (MRO). In parallel, run a project looking at three to six
months of spend. Compare negotiated prices for specific items, prices paid
in other marketplaces and what could have available in Amazon Business or
Work with procurement to establish a basic set of guidelines on when
Amazon Business and Alibaba should be used, considering not only the
price comparisons and lost rebates, but also the less quantifiable factors like
usability and convenience (e.g., time to fulfillment). Also outline when an
alternative buying channel is more appropriate. If possible, embed this logic
into an existing P2P system to route the purchase appropriately.
Collaborate with procurement to embed the buying channel logic into P2P
software. Seek out solutions with a high level of interoperability with these
marketplaces, as opposed to basic integration. For example, future P2P
solutions may have an Amazon-like user experience that can also search
across multiple marketplaces (e.g., Amazon Business, Alibaba,) and online
supplier catalogs (e.g., Staples, Grainger), comparing prices against
company-maintained catalogs and inventory in other locations to make the
best purchasing decision. This is not practical today given some of the
technical and performance limitations of searching vast, external catalogs.
For example, challenges exist like the marketplace vendor's constantly
changing URLs and page layouts to keep out price comparison bots. Future
solutions will begin to address these challenges.
"Willful Disruption: Amazon Disrupts Through Scale, Richness and Reach"
"Vendor Rating: Amazon"
"SWOT: Alibaba Group, Consumer Online Marketplaces, Worldwide"
Strategic Planning Assumption: By 2022, all of the major procure-to-pay
solutions will embed virtual assistants and chatbots for guided buying and self-
Analysis by: Magnus Bergfors and Patrick Connaughton
Over the past decade, cloud-based specialist providers have come to dominate
the P2P space over the native P2P modules of the ERP suites. This has been, in
part, a result of the specialists' superior user experience (UX) and agility, which
has improved end-user adoption and spend under management.
The next wave of improvements in UX lies in leveraging AI technologies such as
virtual employee assistants (VEA) and chatbots. A VEA is form of conversational
agent working on behalf of enterprises to support employee engagement. A
chatbot is also a conversational agent, functionally narrower and often highly
specialized. For more information on VEAs and chatbots, see "Market Insight:
How to Collaborate and Compete in the Emerging VPA, VCA, VEA and Chatbot
The clear use case for a virtual assistant is to actively guide an end user through
the buying process. It asks a series of simple questions to determine the level of
risk associated with the buy and, in turn, the level of control required. The
advantage is that the complexity of business rules determining the best buying
channel and payment method are hidden from the end user. This results in a
more streamlined and compliant process. By interacting with the end user and
also predicting needs, the VEA can create the requisition with the right supplier(s)
and in the right system. Interaction will initially be primarily text — but over time
voice interaction will be a viable option.
Chatbots are already being used by some of the major P2P vendors and the rest
of the competition are close behind, actively piloting the technology. However,
these chatbots are fairly simplistic and used for basic requests, such as finding
specific contracts and POs. Over the next three to five years, we predict that all
leading software vendors will use chatbots to improve the support for buying
organizations. Chat bots will also be used to further automate supplier self-
service inquiry requests. For example, answering common questions like the
status of a payment or to help resolve an order/invoice discrepancy.
As virtual personal assistants (VPAs), the consumer equivalent of the VEA,
mature in the consumer space, business end-user requirements will evolve. The
expectation will be that enterprise applications follow suite. P2P suites that fail to
keep up with evolving end-user expectations will quickly become outdated.
Enterprises that stick with them will struggle with adoption and managing their
A VEA has the potential to serve as the front end for multiple procurement
applications used for different workstreams. These include a vendor
management system (VMS), general-purpose P2P suite or travel and expense
management system. Procurement networks that feature multienterprise grid
capabilities will become central, providing the necessary data to train the VEA.
The more data it has to train on and learn from, the better it will perform. Hence,
data will become an even more critical commodity. There are also other potential
benefits of a procurement network, such as benchmarking and data mining for
Identify specific pain points where a chatbot could quickly add value, like
providing the first line of support for common supplier inquiries.
Work with your current P2P vendor to determine if they have a chatbot
solution that is ready for piloting or in production. If so (and there is a ROI),
create a plan for implementation.
Ensure that the team carefully investigates which languages are supported,
and how well it works outside of demo scenarios in each relevant language.
Design buying processes with a wizard-based user experience in mind.
Embrace the concept of a simple set of questions that leads the end user
down the most compliant, streamlined path. This is in contrast to the
common approach of using a series of intake forms. For example, gathering
all the information at once and applying business rules once the page has
Make sure to start with spend categories where infrequent, casual
requestors are common, because they will benefit the most from a simplified
approach. Once the wizard-based buying approach is in place, VEA rollout
will be faster. This is because the VEAs can leverage the decision tree logic
as a starting point for conversational UI.
CPOs are increasingly seeking new ways to bring value to the business through
innovation. Chatbots and VEAs are perfect examples of the type of innovation
procurement is looking for. Once IT application leaders understand the uses of
chatbots and VEAs, meet with the CPO to explain the technology. Give
procurement the opportunity to endorse the project to who how they are rising to
the challenge of creating a digital business.
"Hype Cycle for Procurement and Sourcing Solutions, 2017"
"Start Preparing Now for the Impact of AI on Procurement"
"Cool Vendors in AI for Conversational Platforms, 2017"
"Magic Quadrant for Procure-to-Pay Suites"
A Look Back
In response to your requests, we are taking a look back at some key predictions
from previous years. We have intentionally selected predictions from opposite
ends of the scale — one where we were wholly or largely on target, as well as
one we missed.
On Target: 2015 Prediction — By the close of 2017, contractual complexity and
audit scrutiny will drive a 40% or higher increase in contract life cycle
management (CLM) solution adoption.
Analysis by: Desere Edwards
CLM adoption is increasing as organizations implement their digital roadmaps to
eliminate paper and improve both their workflows and collaboration. This
prediction "on target" because Gartner's latest forecasts estimate the CLM
market will grow from $267 million spend in 2014, to over $1 billion in 2021. The
market is currently very fragmented with over 160+ vendors. However, there are
clear leaders in revenue and capabilities among the point solution CLM vendors.
Many traditional procurement and sourcing vendors expanded their product
portfolios to offer modules supporting the full source-to-settle (S2S) process,
which includes CLM.
Effective contract management is important for the commercial success of any
business. Compliance concerns and postsignature contract management are
common themes on CLM-related client inquiries. Many organizations use CLM
metadata to track contract terms for audit and regulatory compliance purposes
as a best practice. Several key regulatory, standards and compliance matters
being tracked today by procurement, legal, finance and compliance teams
European General Data Protection Regulation (GDPR): The new data
protection regulation based on privacy risks takes effect in May 2018. This
globally impacts the handling and processing of all personal data on EU
residents. It also strengthens the consent requirement for the processing of
personal data and increases breach notification requirements.
Financial Accounting Standards Board (FASB) and International
Accounting Standards Board (IASB) leasing standards: Accounting
Standards Codification (ASC 842) imposes new standards for leases longer
than one year. One key impact is that lease amounts traditionally classified
as "off-balance sheet" will now be added to the balance sheet. Public
business entities are required to apply the leasing standard for annual
reporting periods beginning after 15 December 2018. Certain not-for-profit
entities and employee benefit plans that file financial statements with the
U.S. Securities and Exchange Commission (SEC) are also subject to the
transition date. All other entities are required to apply the leasing standard
for annual periods beginning after 15 December 2019.
U.S. Treasury Department's Office of Foreign Assets Control (OFAC)
Specially Designated Nationals (SDN) list: The U.S. Department of the
Treasury will enforce penalties on organizations found doing business with
countries and named entities on this list which identifies entities involved in
terrorism, money laundering, drug trafficking, arms dealing and
Physician Payments Sunshine Act (PPSA): The new regulation requires
manufacturers of medical products to publicly report payments made to
teaching hospitals and physicians. It also requires group purchasing
organizations (GPOs) and certain manufacturers to disclose ownership or
investment interests by physicians.
ASC 606 Revenue From Contracts With Customers: New revenue
recognition reporting standards go into effect for public business entities,
certain not-for-profit entities and certain employee benefit plans in December
2017. The new standards apply to all other entities in annual reporting
periods beginning after 15 December 2018.
Missed: 2014 Prediction — By 2016, procurement solutions offering
multienterprise grid functionality will double in number, triggering an order-of-
magnitude improvement in B2B e-commerce agility.
Analysis by: Kaitlynn Sommers
Multienterprise grid functionality has gained traction in the last few years. Vendors
have expanded their capabilities to provide B2B networks that allow a community
of buyers, suppliers and third parties access to common content. This offers
participants the ability to:
Upload sets of data once to exchange on a permission basis.
Publish benchmark data assembled through community transactions.
Plug into a trading community and role-based dashboards within an account
(see "Hype Cycle for Multienterprise Solutions, 2017" ).
Privacy and security have challenged rapid adoption of these tools, which in turn
has limited the order-of-magnitude improvements we predicted. In addition to the
security concerns, user-friendly options like Amazon Business and Alibaba,
which provides instant price comparison for comparable indirect materials, have
gained traction. These options have taking market share from multienterprise
grids provided by traditional P2P vendors. Vendors have also introduced
alternate ways of transacting, such as point-to-point connections, skipping the
grid functionality altogether. For direct materials, business risk remains a concern
for buyers who depend on specific supplier partners to deliver materials with
tailored specifications and quality requirements.
While the availability of multienterprise grid functionality in procurement solutions
has expanded, it has not moved as quickly as we had anticipated. Despite its
challenges, Gartner expects continued adoption to extend over the next five years
leading to improvement in B2B e-commerce agility.