Reasons Given for Manual Alteration

Respondents most often linked manual alteration to unusable data, delayed financial institution decisions, unreliable AI outputs, and problems with third-party feeds and interfaces. The written answers describe reasons for changing or completing the work by hand. Table 5.7 maps those descriptions to the relevant survey questions and rated indicators. These mappings connect the accounts to Part 2 without establishing that the reported reasons caused the study outcomes.

1 Reasons for Manual Alteration of Financial-Institution-Provided Data

Data quality was the most frequently reported reason for manual alteration of financial-institution-provided data. It was identified in 341 of the 588 method-based written answers, covering 103 engagements. Borrowed and shared credentials were identified in 92 answers and normalization across source systems in 89. The accounts connected recurring correction with data quality and restricted access to the systems in which the problems originated.

Recurring correction in these accounts concerned the suitability of inputs for agent use and access to the systems in which errors originated. Correcting a supplied file does not by itself resolve the source-data or access problem. The reported reasons connect manual alteration to the data quality, accessibility, and normalization indicators examined in Part 2, “Financial-Institution-Provided Data.”

“I re-key the SSI details every time the layout shifts between deliveries, we still don't have working access to the source, so it's manual. Every time.” -Individual Contributor, BPO, Trade, settlement and reconciliation

“We scrub the incoming Data clean every single cycle, as well as that it just turns up dirty again the very next time, so I just spend hours rebuilding same fields week after week” -Individual Contributor, ITO, Technology engineering and operations

“Credentials to the client Repository expire and nobody renews them, so I use a workaround login just to keep the files moving… it's the same thing every status call, we chase it and nothing changes” -Individual Contributor, BPO, Corporate and back-office functions

i cleaned the policy data again before load, we clean it every cycle and it comes back dirty again, i still don't have proper access to the source system so i pull what i can and patch the rest by hand.”

INDIVIDUAL CONTRIBUTOR · ITO, Client servicing and communications

2 Reasons for Manual Alteration of Financial Institution Processes

Delayed decisions were the most frequently reported reason for process manual alteration. The financial institution’s speed of response was identified in 229 of the 857 method-based written answers, covering 108 engagements. Process documentation was identified in 138 answers and financial institution management and decision authority in 94. These accounts describe delivery teams changing the work when the decisions or documented procedures needed to perform it were unavailable.

A changed process may allow work to continue while leaving its authoritative definition or the required financial institution decision unresolved. These accounts connect manual alteration to process ownership and authorization, not simply to the frequency of intervention. Part 2 distinguishes process documentation from the financial institution's decision authority and response speed. The accounts here do not establish that every process change lacked approval or was improper.

“Process never written down anywhere, everyone do it different way. I ask questions and nothing come back so I change step myself to keep work moving.” -Individual Contributor, CXM, Corporate and back-office functions

“What they gave us was a slide, not a procedure. And change control cant keep up with the speed we work at so I just adjust it” -Individual Contributor, CXM, Corporate and back-office functions

“Last month the Client raise a request that need urgent proccess change but formal Change board only meet monthly, so I make the update myself and inform team lead after” -Individual Contributor, KPO, Credit, underwriting and suitability

“The change process was built for quarterly release, not for agent making decision every day… so when servicing rule need small update I please make the same myself rather than wait weeks for formal sign off” -Individual Contributor, BPO, Payments, cards and disputes

3 Reasons for Manual Alteration of AI Outputs

Unreliable output before human review was the most frequently reported reason for AI-output manual alteration. It was identified in 321 of the 429 method-based written answers, covering 106 engagements. Inconsistency across repeated runs was identified in 176 answers. The separate reasons compilation recorded instructions to continue processing outputs without raising the problem in 16 answers from 12 respondents on 10 engagements. This differs from the method-based count of 14 answers on nine engagements reported earlier. The counts remain separate because the coverage of that theme has not been reconciled.

The reported correction and selection of outputs place human review within the work required to produce an acceptable result. A corrected output cannot by itself demonstrate the factual reliability or repeatability of the agent's uncorrected work. Whether the human contribution was recorded and visible in performance reporting is examined separately below. The two escalation counts remain descriptions from their respective compilations.

“yesterday i ran same question twice and bot gave different policy answer both time, nothing else changed. and i had to correct it before new hire saw wrong one.” -Individual Contributor, CXM, Corporate and back-office functions

“Nothing about it is repeatable, i get different Output on same Input twice in one day and Change process assumes normal release cycle not agent running every day” -Individual Contributor, KPO, Trade, settlement and reconciliation

“The tool gives a different attribution breakdown each time I rerun it… consistency is the one thing it cannot do and my own leadership does not want that raised with the client” -Individual Contributor, KPO, Trade, settlement and reconciliation

“It changes its mind between one pass and the next on the same input, this has occured on nearly every run this week, pressure from our own side is only to keep it moving, so I just pick the best version and send” -Individual Contributor, ITO, Technology engineering and operations

4 Reasons for Reporting Alteration

Factual reliability before human review was the most frequently identified additional subject in accounts of reporting alteration, appearing in 53 answers on 37 engagements. Most accounts described how reports were altered rather than supplying a separate reason for the practice. Table 5.2 distinguishes those reporting methods from the additional survey subjects summarized here.

The distinction between a reporting method and a reason for using it limits interpretation of these counts. Accounts of corrected work attributed to agents connect reporting alteration to the visibility of human effort, but identifying such a method does not independently establish the respondent's or vendor's motive. The relevant comparison is between the work described and what its reported performance represented.

“The vendor's onboarding metrics counted a task as agent-handled even when a human quietly redid it afterward because the agent's output was unusable, which overstated automation coverage.” -Individual Contributor, CXM, Corporate and back-office functions

The vendor's delivery dashboard reported reporting-rule coverage as a percentage of rules migrated into the agent, but counted a rule as migrated the moment a draft version was uploaded, regardless of whether an SME had validated it, so coverage looked near complete while unvalidated rules kept producing quiet errors downstream.”

INDIVIDUAL CONTRIBUTOR · KPO, Regulatory, audit, tax and legal

5 Reasons for Third-Party Work Done by Hand

Unusable feeds, unstable interfaces, and insufficient notice of provider changes were the reported reasons for third-party work done by hand. Feed-data usability was identified in 36 of the 113 method-based written answers, covering 27 engagements. Interface and data-structure stability and advance notice from third-party providers were each identified in 29 answers. The accounts describe manual work needed when the delivery team could not rely on the third-party connection or its data.

These accounts place part of the effort needed to use third-party services within the delivery team's own work. They identify a potential continuity dependency on the people who know how to complete that work by hand. The related ratings of retained manual capability and replacement of informal corrections remain separate from counts of reported occurrence.

“Their bureau feed changed a field layout back in march and nobody flagged it… we found it when the challenger run wouldn't reconcile. Two days gone. And half of what we need from the Bureau portal has no real interface. You log in and export it by hand. That part is invisible here.” -Individual Contributor, KPO, Credit, underwriting and suitability

“i type the return into Submission portal by hand every time, agent cannot use a screen made for a person. give us a proper interface, and please tell us before they change the format, that is what has to happen” -Individual Contributor, KPO, Regulatory, audit, tax and legal

The feeds themselves have settled — the supplier registry and sanctions data have not moved under us once this year, and our contracts with those providers now say something sensible about notice periods, which was not true in January. Where we have not moved is our own map of who sits behind them. We know our direct providers; we do not honestly know their sub processors, and the truthful answer on the fourth party question is that we never finished the exercise. Nor could we run the onboarding checks by hand for long if one of them went dark — the people who used to do that work are doing something else now.”

ENGAGEMENT MANAGER · BPO, Corporate and back-office functions · “

The vendor's dashboard reported average turnaround time per valuation without disclosing that a meaningful share of deliverables were quietly redone by senior staff after bankers rejected the initial agent output, understating real cycle time.”

INDIVIDUAL CONTRIBUTOR

KPO, Corporate and back-office functions