Vendor Ecosystem Restructuring Agenda

From WFM Labs


Restructuring a vendor ecosystem is the set of structural decisions a client makes when it takes over, consolidates or renegotiates a group of outsourced service suppliers as one system rather than as separate contracts. The decisions fall into seven groups: what the suppliers are for, how many there are and how they are partitioned, which work goes to them, what unit the client pays for, what the contract terms protect, how performance is measured, and who governs the relationship day to day. Almost every lever in the catalog depends on one that sits beneath it. The dependency order therefore constrains the sequence in which levers can be decided. Pricing changes depend on measurement; measurement depends on a shared definition of the work; the definition of the work depends on consolidation having made a single instrument affordable; and everything depends on an explicit statement of what a vendor is for.

This page states the agenda first as a plain list, then gives each item its reason, its dependency, and what the research and this wiki's own doctrine say about it. A leader examining any service business with an outsourced supply base can select from it to build a short list of near-term objectives. It does not recommend a particular set of choices, because the right set depends on the estate.

The restructuring ladder. Consolidation, one measurement instrument, one work taxonomy and a contract inventory come first and can run in parallel. The comparability gate sits above them: nothing higher on the ladder is decided on data until the suppliers can be compared on the same instrument and the same definition of the work. Governance placement is not a rung; it runs alongside every tier.

The agenda in one list

The structural changes a leader examining any outsourced service supply base would put on the table, in dependency order. Each item is expanded in the tiers that follow.

  • State what each supplier is for. Variance absorption or market access; a seat that can name neither is a consolidation candidate. Treat the back office as the standardize-consolidate-automate track, not the valve.
  • Consolidate by body of work. A pair per body, front office and back office, chosen on different logic: substitutability for the front, automation path for the back.
  • Bring every placement on-network and under contract. Inventory supplier-by-site-by-system first; exit live-voice work the client cannot see; tolerate asynchronous work off-network last.
  • Build one measurement instrument per body of work and write it into every contract. Change the instrument rather than sampling harder. Set targets by work tier, never by location.
  • Define the work once. One taxonomy across suppliers and captives, as the entry ticket for transaction pricing, automation and any placement platform.
  • Read every contract. Term, notice, minimums, attrition and tenure terms, at-risk fees, facility approvals, data and automation clauses, from the signed documents, largest supplier first.
  • Move the pricing unit up the ladder. Clock hour to productive hour to transaction to outcome, standardizing the productive-hour denominator before it ranks anything.
  • Contract for tenure by team. Attrition ceiling on the named account, tenure-mix floor, named-team continuity. Keep job design with the operational line.
  • Put fee at risk on the one scorecard that also governs allocation, framed as earnable, settled monthly.
  • Set commitment horizons against forecast error with the planning function in the room and a lock ladder of tolerances.
  • Name the relief valve in the contract: committed versus contingent flex, metered, with the ramp-down protected and cross-supplier overflow tested.
  • Design renegotiation at signing. Flexibility provisions and termination for convenience, so amendments improve outcomes rather than invite rent seeking.
  • Decide dedication explicitly. Price the seats that are genuinely dedicated; pool the rest; offer a discount for unconstrained delivery.
  • Pilot transaction pricing on one back-office transaction type once the taxonomy and instrument exist.
  • Share the automation gain. Gain-share on the efficiency component; data isolation, no-training and model-ownership clauses; automation-driven volume decline treated explicitly against minimums. Draft now, sign at the next renewal or migration.
  • Split the commercial pen from the daily performance conversation. Vendor team holds contracts, billing, penalties and the instrument; the operational line holds daily performance for all but highly specified work; the instrument owner never owns the verdict.
  • Run governance in tiers with a cross-supplier calibration forum.
  • Plan and run exits with a notice-and-term inventory, parallel run, hypercare and knowledge transfer into the taxonomy.
  • Build toward a placement function that computes placement rather than deciding it, with the consolidated pair as its first supply-side input and flexibility as a measured objective.

Tier 0: what a vendor is for

Every lever below inherits its logic from a statement of purpose, and a supply base that cannot state one will be restructured on price by default. This wiki's sourcing doctrine holds that an outsourced seat should carry one of two justifications: variance absorption, meaning the seat exists to take volatility that fixed captive structures should not carry, or market access, meaning the seat exists in a labor market or language the client cannot serve captively (see Outsourcing as a Risk Lever and The Vendor as Relief Valve). Where the client already operates a low-cost captive footprint, the rate argument for outsourcing largely disappears, and a seat that can name neither justification is a candidate for consolidation rather than renegotiation.

The purpose statement changes the meaning of every commercial lever. A supplier bought for variance absorption is priced on flexibility, and its contract should protect the release direction of the ramp. A supplier bought for market access is a fixed node, and its contract should protect continuity and tenure. A supplier bought for rate arbitrage is priced on the unit cost of an hour, and its contract should protect the hour. Restructuring a portfolio without deciding which of these each supplier is produces terms that protect the wrong thing.

The back office is the exception that proves the rule. Back-office work absorbs variance through its backlog at no premium, so the variance-absorption justification does not apply to it. Its justification is cost and automation path: it is the first body of work to standardize, consolidate and automate (see Consolidate, Then Automate: Back-Office Fulfillment).

Tier 1: portfolio structure

Lever What changes Why Depends on What the evidence says
Consolidate by body of work The roster shrinks to a small number of suppliers per body of work, typically a pair for front office and a pair for back office Fewer relationships to govern; in this wiki's doctrine, consolidation is what makes one calibrated measurement instrument affordable Tier 0. Consolidation is the one lever that is not gated on anything else Dyadic contracting theory falls short for multisourcing (Bapna et al. 2010); a supplier portfolio can emerge from combined top-down and bottom-up decisions rather than one design (Levina & Su 2008); cloud services are less often multisourced and supplier experience matters (Handley et al. 2022). Detail below
Two per body, not one or three Each body of work keeps a comparator and a continuity option One supplier has no comparator; three splits volume until each tranche is too small to measure and too small to matter Consolidation This wiki's design doctrine (Vendor Portfolio Design by Body of Work). The literature on the optimal count is thin, and the page says so
A different selection logic per body Front-office suppliers are chosen on shared properties (language, breadth, footprint, scale) so they can substitute for each other; back-office suppliers are chosen on automation path, with at least one able to automate what its own agents handle Front office is bought for continuity and competition; back office is bought for its trajectory toward not needing agents Tier 0; the front/back distinction Front-Office and Back-Office Outsourcing Are Different Bodies of Work
Eliminate off-network and uncontracted work Work delivered on supplier-owned telephony, routing or tooling that the client cannot see, or work placed with a supplier outside any framework agreement, is brought on-network or exited Off-network delivery defeats comparability and produces per-system quality workarounds; uncontracted placement is procurement non-compliance A supplier-by-site-by-system inventory, which most estates do not hold Non-compliant purchasing ranges from unintentional to deliberate and has identifiable organizational causes (Karjalainen, Kemppainen & van Raaij 2009). Whether live voice can ever be governed off-network is a judgment, not a finding; asynchronous work tolerates it better
Exit and transition governance Suppliers consolidated out get a notice-and-term inventory, a parallel run, hypercare, closure and knowledge transfer into the shared taxonomy A recurring failure in the outsourcing literature is underestimated transition and monitoring cost The contract inventory Hidden costs cluster in search and contracting, transition, managing the effort, and post-outsourcing transition (Barthélemy 2001). Detail below

The multisourcing evidence is thinner than the consolidation lever deserves. Bapna, Barua, Mani and Mehra (2010) argue that extrapolating dyadic contracting theory to multiple competing suppliers falls short, and set out a research agenda rather than results. In a longitudinal case of one financial institution, the supplier portfolio emerged from combined top-down strategy and bottom-up middle-management decisions rather than from a single design (Levina & Su 2008). Cloud-based services are less likely to be multisourced, automation's effect is context-specific, and experiential learning with a supplier matters (Handley et al. 2022). On exits, this wiki's doctrine holds that transition cost rather than gross benefit is usually the binding variable, so substitution is best applied at new business, renewal and attrition replacement rather than as a relocation program.

The inventories that the last two levers depend on, a site-by-system inventory and a contract inventory, are Tier 2 artifacts. Tier 1 and Tier 2 are therefore parallel workstreams rather than strict steps, and the numbering records the order in which decisions are taken, not the order in which the supporting work is finished.

Tier 2: the comparability foundation

These three levers run in parallel and gate everything above them. Until they land, a comparison between two front-office suppliers is a comparison of noise, and a comparison between two back-office suppliers is a comparison of different work.

One measurement instrument per body of work. One rubric, one sampling design, the same coverage and the same calibration cadence across every supplier and every captive center doing the same body of work. The resolution of a quality comparison is set by the absolute number of observations, not by the sampling percentage, and manual sampling at typical rates cannot detect the differences that allocation decisions turn on (see Sample Size and Detectable Difference in Quality Measurement). The fix is to change the instrument, not to sample harder, and that is a contracting decision rather than a quality-team decision, because the instrument has to be written into every supplier's terms. Targets are set by work tier, never by location or supply type: a location-differentiated target announces that delivery from one place is expected to be worse and degrades the aggregate mechanically as that location's share grows.

One work-type taxonomy. The same work has to be defined before it can be compared, priced per transaction, or automated. A shared taxonomy is also the entry ticket for any placement or automation platform, because whatever the platform is fed, it locks in (see WFM Data Governance and Quality and Service Chain Decomposition and Node Sourcing). Contract research finds that detailed task description and contingency planning behave as complements and that both improve with repeated contracting (Argyres, Bercovitz & Mayer 2007), so the taxonomy pays back in contract quality as well as in measurement.

A contract inventory. Term, notice, minimum volumes, attrition ceilings, tenure floors, named-team continuity, at-risk fees, location and facility approvals, automation and data clauses, for every supplier, read from the signed documents rather than from institutional memory. In practice the largest supplier is often the one with the least on file. The inventory runs in parallel with the other two because renewal calendars do not wait for measurement.

Tier 3: pricing unit and terms

The pricing-unit ladder

The unit a client pays for determines what the supplier optimizes. The units form a ladder, and each step transfers a risk from client to supplier.

Unit What the client pays for Risk transferred to the supplier What it requires
Full-time equivalent or clock hour Time on the roster None; the client bears shrinkage, absence and idle time Nothing beyond a headcount report
Productive hour Time in a productive state Shrinkage: breaks, training, absence, and the supplier's own scheduling inefficiency An agreed definition of a productive hour and an auditable state model
Transaction Units of work completed to standard Handle-time variance and the productivity of the supplier's own people One taxonomy of transactions, a quality standard per transaction, and a volume commitment that survives automation
Outcome A business result (resolution, retention, sale, satisfaction) Attribution and baseline disputes; the supplier's control over factors outside its process An attribution model both parties accept, and a measurement instrument the client owns

Contract-choice research in offshore services finds that requirement uncertainty, team size and resource shortage predict whether a supplier will accept fixed-price or time-based terms, and that contract choice significantly determines supplier profit (Gopal, Sivaramakrishnan, Krishnan & Mukhopadhyay 2003). In contact-center contracting specifically, pay-per-time, pay-per-call and service-level terms coordinate the supplier's capacity decision differently under information asymmetry about agent productivity, and no single form dominates (Hasija, Pinker & Shumsky 2008; Akşin, de Véricourt & Karaesmen 2008; Ren & Zhou 2008). Moving up the ladder is a series of definitional projects, not a single commercial negotiation.

The productive-hour denominator has to be standardized before it is used to rank anything. Productive hours per full-time equivalent vary materially between locations and suppliers, and a rate per productive hour divides by that number. Where a share of the spread between locations is definitional, the location ranking moves, and moves most at the extremes. A client that shifts to productive-hour pricing without fixing the denominator has changed what it pays for without knowing what it is paying.

Terms that price the right thing

Tenure by team. Where a supplier is bought for market access or for judgment-heavy work, the client is buying a stock of accumulated experience. The stock cannot be purchased directly, but its drain rate can be constrained through three terms: an attrition ceiling on the named account rather than the site, a tenure-mix floor expressed as the share of the named team above a stated tenure rather than as an average, and named-team continuity (see Placement Rules and the Tenure Contract and Speed to Proficiency Curve). In the operator judgment behind this page, named-team continuity is the term suppliers resist hardest and the one that matters most. Job design is a stronger lever on turnover than any contract term available here, and it sits with the operational line rather than in the contract: turnover in service work ran near nine percent where jobs were high-discretion and low-monitoring and near thirty-six percent where the reverse held (Holman, Batt & Holtgrewe 2007).

An at-risk fee on the same scorecard. In this wiki's allocation doctrine, volume reallocation between suppliers is a weak incentive on its own, because the contestable tranche is small and the response is slow. A share of fee at risk, settled monthly on the same scorecard that governs allocation, is the stronger instrument (see Performance-Based Vendor Allocation Design). Two scorecards produce conflicting supplier behavior, and the doctrine treats a second scorecard as the structural defect to look for first. Penalty framing raises compliance effort and lowers knowledge-sharing and affective commitment (Fehrenbacher & Wiener 2019), so the at-risk share should be framed as an earnable bonus where the client wants the supplier to share what it learns.

Commitment horizon set against forecast error. Volume commitments are set commercially, are meetable only operationally, and are justified by a forecast produced by a planning function that is often not in the room. Comparisons of forecasting methods for intraday arrivals find that accuracy degrades quickly with lead time and that simple benchmarks are hard to beat (Taylor 2008), and forecast error exceeds the stochastic variance that staffing models assume (Steckley, Henderson & Mehrotra 2009). No published work ties forecast error by horizon to a defensible commitment horizon; the gap is a named absence in the literature. The practical instrument this wiki's doctrine proposes is a lock ladder: a long outlook that is not a commitment, a medium-horizon forecast locked unless it moves beyond a tolerance, and progressively tighter interval-level tolerances close to the day, with a billing cap on over-lock. Quantity-flexibility and option contracts from supply-chain theory transfer in shape but not in their efficiency proofs, because labor has no storability and no salvage value.

Committed and contingent flex. A supplier bought for variance absorption needs its valve named in the contract, with a distinction between committed flex, meaning scheduled capacity paid daily, and contingent flex, meaning a retainer for readiness and full rate on call. The release direction is the half that matters most: ramping the volatility book down without severance, stranded idle time or burned tenure. A relief valve that is open all the time is a leak, so the valve is metered (see The Vendor as Relief Valve and Chaining and Flexibility Design).

Renegotiation designed in advance. Flexibility provisions, termination-for-convenience rights and supplier reuse rights are associated with amendments that improve outcomes for both parties rather than with rent seeking (Susarla 2012); the author's interpretation is that renegotiation design at signing is what makes the difference. Formal contracts and relational governance work as complements (Poppo & Zenger 2002), so a restructuring that hardens terms without building the relationship that interprets them tends to produce the gaming that service-level contracts invite (Milner & Olsen 2008; Holmström & Milgrom 1991).

Dedication, designation and pooling

In this wiki's doctrine, dedicated capacity is a product feature with a price, and it is frequently supplied without a corresponding price. Estate reviews behind the doctrine found that much of the capacity described as designated was designated on paper and shared in operation, and that designation drifts because it is emergent rather than designed. The commercially cleaner inversion is to offer a discount where the client permits unconstrained delivery. Splitting a pooled queue carries a measurable pooling penalty, and a round-the-clock language commitment carries a fixed staffing cost before absence cover regardless of volume (see Sourcing Design Axes: Node and Client Ownership and Comparing Delivery Arrangements). A restructuring should decide which seats are genuinely dedicated, price them as such, and let the rest pool.

Tier 4: automation and who captures the gain

Automation changes the supply base in three ways that the contract has to anticipate.

It changes what the supplier is paid to suppress. A supplier paid per hour or per transaction is paid more when volume rises, so the largest available efficiency lever, contact reduction, is one the supplier is structurally disincentivized from helping with. A gain-share against the efficiency component realigns this. Practitioner accounts report first-year returns above thirty percent from service automation (Lacity & Willcocks 2016); the figure is self-reported and not independently verified, and the gain-share logic does not depend on it.

It creates ownership questions that most contracts do not address. Who owns a model that learned on the client's transaction data, whether the supplier may train on that data, how automation-driven volume decline is treated against minimum commitments, and how a productivity dividend is shared. In the operator experience behind this page, these clauses are among the most frequent redlines in current renewals and are white space in most existing agreements. They are timed by the counterparty's calendar, meaning any platform migration or renewal, rather than by the client's ladder.

It changes the mix of what remains. Scripted contacts are automated and deflected first, so the human queue drifts toward exceptions. That moves the work along the specification gradient and moves day-to-day oversight of it toward the operational line (see Specification and the Placement of Vendor Oversight). A restructuring that fixes governance placement before automation has to expect to revisit it.

Tier 5: governance

Split the commercial pen from the daily performance conversation. The shared-service vendor team holds contracts, billing, penalty enforcement, formal quality call-outs and the measurement instrument; the operational line that designs the job holds daily performance, coaching and procedural adherence for all but the most highly specified work; ramps are run jointly; commitment setting includes the planning function (see Vendor Governance Placement). Enforcement is separated from relationship management because penalty framing and knowledge-sharing pull in opposite directions.

The row that must not collapse is measurement. The same instrument, across suppliers, captive centers and in-market delivery, owned by a party that does not own the verdict. The measured party never controls measurement, and that rule applies to the client's own operational line as much as to the supplier. The gaming premise is well evidenced (Ridgway 1956; Bevan & Hood 2006); the ownership rule itself is this wiki's doctrine.

Tiered cadence. Strategic, tactical and transactional reviews at different intensities, with an allocation body, a model-review body, a monthly business review and a cross-supplier calibration forum. Reviews of centralized versus decentralized procurement report that supplier evaluation systems are weak or absent without a central program (Kanepejs & Kirikova 2018), which is the argument for the shared-service vendor team holding the instrument even where it does not hold the daily conversation.

Tier 6: the destination

The tiers below feed a placement function that computes where work goes rather than deciding it, holding what produces the answer rather than the answer itself, with the consolidated front-office pair as its first supply-side input (see Placement Engine Architecture and Integrated Resource Optimization Center). Flexibility is promoted from a priced contract term to a measured objective, because rigidity is invisible in steady state and a model with no flexibility term recommends exactly the estate that cannot respond. The decisions in Tiers 1 to 5 cannot wait for the engine, and the engine cannot be built without the decisions' data, so the two tracks run in parallel rather than in sequence.

Dependencies, and the order they impose

  1. Tier 0 is stated first and costs nothing but a decision. Every supplier is assigned a justification.
  2. Consolidation is not gated on anything. It is the decide-now lever, and its exit sequencing follows the client's own site plans.
  3. The step after consolidation is not "allocate on performance." It is "make the suppliers comparable." One instrument and one taxonomy run first and in parallel.
  4. The contract inventory runs in parallel too, because renewal calendars set their own deadlines, and the supplier with the least on file goes first.
  5. Tenure terms, productive-hour pricing and the at-risk fee can be written once the inventory shows what is already promised, and once the productive-hour denominator is standardized.
  6. Transaction and outcome pricing, automation gain-share and the relief-valve test follow the comparability gate. A transaction price on an undefined transaction is a rate card with a new name.
  7. Automation and model-ownership clauses are timed by the counterparty's migrations and renewals, not by the ladder. They are drafted early and held ready.
  8. The placement engine takes the consolidated pair as its first input and is built in parallel with, not after, the decisions.

A one-page objectives template

A leader building a near-term objectives page from this catalog can use the following frame. Each row is a lever; the columns are the decision it requires, the gate it sits behind, the kind of owner, the horizon or trigger, and the condition that evidences completion.

Lever Decision required Gate Owner type Horizon or trigger Done when
Purpose per supplier Which of the two justifications each seat carries None Sourcing leadership Now Every seat carries one justification in the sourcing register
Consolidate by body of work Which pair per body; which suppliers exit and on what sequence None Sourcing leadership with the vendor team Now Target roster and exit sequence approved
Off-network and uncontracted work Inventory first; then bring on-network or exit, with asynchronous work last Site-by-system inventory Vendor team with technology Now to next quarter No live-voice work off-network; all placements under a framework agreement
Exit and transition governance Notice and term inventory, parallel run, hypercare, closure, knowledge transfer Contract inventory Vendor team with operations Per exit Each exit closed with knowledge captured in the taxonomy
One instrument Rubric, sampling design, calibration cadence written into every contract None; gates Tiers 3 to 6 Vendor team owns the instrument; quality runs it Now One rubric and sampling design in force across all suppliers and captives per body
One taxonomy Definition of the same work across suppliers and captives None; gates transaction pricing and automation Operations with planning Now Shared taxonomy published and used in reporting and contracts
Contract inventory Every term read from signed documents; largest supplier first None Vendor team Now Every supplier's terms on file, from the signed documents
Productive-hour pricing Standardize the denominator; extend the unit to remaining agreements Denominator standard Vendor team with planning Next quarter One denominator definition; all agreements on productive hours
Tenure by team Attrition ceiling on the account, tenure-mix floor, named-team continuity Contract inventory Vendor team; job design stays with operations Next quarter Three terms contracted and reported monthly
At-risk fee on one scorecard Share at risk; monthly settlement; bonus framing One instrument Vendor team Next quarter One scorecard governs both fee and allocation
Commitment horizon Lock ladder with tolerances; planning function in the room Forecast-error profile by horizon Planning with the vendor team Next two quarters Lock ladder in every agreement; planning signs every commitment
Transaction pricing for back office Pilot on one transaction type Taxonomy and instrument Vendor team with operations Next two quarters One transaction type priced and settled per unit for a full quarter
Automation gain-share and model clauses Gain-share on the efficiency component; data and model ownership; volume decline against minimums None to draft; counterparty calendar to sign Vendor team with legal and technology Trigger: next renewal or platform migration Clauses drafted now; signed at the trigger
Relief-valve terms and test Committed versus contingent flex; metering; cross-supplier overflow tested annually Purpose per supplier Planning with the vendor team Next two quarters Valve named in contract, metered, and tested once
Dedication and pooling Which seats are genuinely dedicated; price them; pool the rest Contract inventory Vendor team with operations Next quarter Every seat classified and priced to its class
Governance placement Commercial pen with the vendor team; daily performance with the line; measurement row protected None Vendor team and operations jointly Now Responsibilities documented and the instrument owner is not the verdict owner
Placement engine First supply-side input; flexibility as a measured objective Taxonomy Planning Trigger: runs alongside all tiers Consolidated pair loaded as the first supply-side input

Where the evidence runs out

The catalog combines findings with doctrine, and the page should be read with the distinction in view. The control-mode, contract-choice, multisourcing and hidden-cost findings are peer-reviewed. The following are this wiki's design doctrine, consistent with those findings but not reported by any study: the two-justification rule for what a vendor is for; two suppliers per body of work; the claim that consolidation is what makes one instrument affordable; the weakness of volume reallocation as an incentive and the second scorecard as the defect to look for; named-team continuity as the term that matters most; tenure-mix floors rather than average tenure; the tenure-matched same-work comparison; the lock ladder as the commitment instrument; the observation that designated capacity is often shared in operation and dedication often unpriced; the claim that transition cost is the binding variable; the governance split and the rule that the instrument owner does not own the verdict; the cross-supplier overflow clause; and the correlated-surge failure of shared vendor capacity. The claim that automation and model-ownership clauses are among the most frequent current redlines is operator experience, not a market study. The commitment-horizon problem is an absence in the literature rather than a result. The off-network exit criterion is operator judgment imported from prior estates and should be treated as such until an estate's own inventory supports it.

Maturity Model considerations

At L1–L2, suppliers are managed as separate contracts, each on its own scorecard and pricing unit, and consolidation happens by attrition rather than design. At L3, the portfolio is consolidated and a shared scorecard exists, but the pricing unit is still time and the taxonomy is still the supplier's. At L4, one instrument and one taxonomy govern every supplier and captive center, pricing is on productive hours or transactions with the denominator standardized, tenure terms are contracted, and governance placement follows the specification of the work. At L5, commitments are set against a stated forecast-error budget, flexibility is a measured objective, automation gain-share is contracted, and placement is computed by a function that sits outside the structures it allocates to.

See Also

References

  • Akşin, Z., de Véricourt, F. & Karaesmen, F. (2008). Call center outsourcing contract analysis and choice. Management Science, 54(2).
  • Argyres, N.S., Bercovitz, J. & Mayer, K.J. (2007). Complementarity and evolution of contractual provisions: an empirical study of IT services contracts. Organization Science, 18(1). 386 contracts.
  • Bapna, R., Barua, A., Mani, D. & Mehra, A. (2010). Cooperation, coordination, and governance in multisourcing: an agenda for analytical and empirical research. Information Systems Research, 21(4).
  • Barthélemy, J. (2001). The hidden costs of IT outsourcing. MIT Sloan Management Review, 42(3).
  • Bevan, G. & Hood, C. (2006). What's measured is what matters: targets and gaming in the English public health care system. Public Administration, 84(3).
  • Fehrenbacher, D.D. & Wiener, M. (2019). The dual role of penalty: the effects of IT outsourcing contract framing on knowledge-sharing willingness and commitment. Decision Support Systems, 121.
  • Gopal, A., Sivaramakrishnan, K., Krishnan, M.S. & Mukhopadhyay, T. (2003). Contracts in offshore software development: an empirical analysis. Management Science, 49(12). 93 projects.
  • Handley, S.M. et al. (2022). The single-sourcing versus multisourcing decision in information technology outsourcing. Journal of Operations Management, 68(6–7).
  • Hasija, S., Pinker, E.J. & Shumsky, R.A. (2008). Call center outsourcing contracts under information asymmetry. Management Science, 54(4).
  • Holman, D., Batt, R. & Holtgrewe, U. (2007). The Global Call Center Report. ~2,500 centers, 17 countries.
  • Holmström, B. & Milgrom, P. (1991). Multitask principal-agent analyses: incentive contracts, asset ownership, and job design. Journal of Law, Economics, and Organization, 7.
  • Kanepejs, E. & Kirikova, M. (2018). Centralized vs. decentralized procurement: a literature review. CEUR Workshop Proceedings, 2218.
  • Karjalainen, K., Kemppainen, K. & van Raaij, E.M. (2009). Non-compliant work behaviour in purchasing: an exploration of reasons behind maverick buying. Journal of Business Ethics, 85(2).
  • Lacity, M.C. & Willcocks, L.P. (2016). A new approach to automating services. MIT Sloan Management Review, 58(1).
  • Levina, N. & Su, N. (2008). Global multisourcing strategy: the emergence of a supplier portfolio in services offshoring. Decision Sciences, 39(3).
  • Milner, J.M. & Olsen, T.L. (2008). Service-level agreements in call centers: perils and prescriptions. Management Science, 54(2).
  • Poppo, L. & Zenger, T. (2002). Do formal contracts and relational governance function as substitutes or complements? Strategic Management Journal, 23(8).
  • Ren, Z.J. & Zhou, Y.-P. (2008). Call center outsourcing: coordinating staffing level and service quality. Management Science, 54(2).
  • Ridgway, V.F. (1956). Dysfunctional consequences of performance measurements. Administrative Science Quarterly, 1(2).
  • Steckley, S.G., Henderson, S.G. & Mehrotra, V. (2009). Forecast errors in service systems. Probability in the Engineering and Informational Sciences, 23(2).
  • Susarla, A. (2012). Contractual flexibility, rent seeking, and renegotiation design: an empirical analysis of information technology outsourcing contracts. Management Science, 58(7). 141 contracts.
  • Taylor, J.W. (2008). A comparison of univariate time series methods for forecasting intraday arrivals. Management Science, 54(2).
Retrieved from "https://wiki.wfmlabs.org/w/index.php?title=Vendor_Ecosystem_Restructuring_Agenda&oldid=4956"