Applied Human Conduct Review — Issue No. 0009
This issue asks how a community's values get written down, who writes them, and how they are kept from hardening. Four brief reports follow member-facing teams giving an AI assistant situation-specific notes on what their community expects, customer-facing sites using an AI interviewer to draw out what frontline staff know about treating customers, hiring panels that screen for conduct and then prefer candidates unlike the existing team, and teams that answer what their newcomers notice. The values that held up best were set by the people closest to the work, and they were kept open to people who saw things differently. Making them explicit took someone's time: staff writing notes, veterans in editing sessions, peers paired with new hires, leads answering questions. As with all issues of Applied Human Conduct Review, reports favor concrete, replicable behavioral comparisons over normative argument.
Papers in This Issue
In 36 member-facing teams across eleven community organizations, all using the same AI assistant to draft written replies, we randomly assigned teams within each organization to one of three set-ups for twelve weeks. In "value-note" teams (n=12), staff wrote short notes on what their community values in specific situations, and the assistant drew on the two or three notes matching each incoming message. "Values-prompt" teams (n=12) used one fixed paragraph describing the organization's values on every request, and "default" teams (n=12) used the assistant as supplied. Members of each community, blind to set-up, rated sent replies from value-note teams 5.3/7 for fitting "how we treat each other here," against 4.5/7 for values-prompt teams (an 18% gap) and 4.2/7 for default teams (a 26% gap). Among drafts sent without substantial staff edits, the ratings were 5.2/7, 4.3/7 (a 21% gap), and 4.1/7 (a 27% gap). Follow-up complaints or escalations within 14 days ran at 3.4 per 100 replies for value-note teams, against 4.2 (a 24% gap) and 4.4 (a 29% gap). The notes cost staff a median 34 hours per team in the first four weeks and 1.6 hours a week after that. By week twelve, 27% of incoming messages in value-note teams still matched no note, and replies to those messages were rated 4.6/7, about the same as values-prompt replies.
In 42 customer-facing sites across six organizations, we randomly assigned sites within each organization to one of three ways of writing down how their staff treat customers. At "AI-interview" sites (n=14), an AI interviewer asked frontline staff to recount specific customer interactions that had gone notably well or badly, probed for what they did and why, and grouped the accounts into a draft playbook that a small staff group edited in one two-hour session. At "workshop" sites (n=14), a human facilitator ran two half-day workshops producing the same kind of playbook, and "head-office" sites (n=14) kept their existing service guidelines. More staff contributed at AI-interview sites (66%) than attended workshops (52%), and AI-interview playbooks described more distinct situations (46 versus 37, a 24% gap). But fewer of their entries said when the usual practice should be set aside (21% versus 27% for workshop playbooks). Over their first three months, the 318 new hires at these sites received customer ratings of 5.5/7 at AI-interview sites, 5.6/7 at workshop sites, and 4.8/7 at head-office sites (gaps of 15% and 17% over head-office sites). In interactions involving a problem, such as a disputed charge or a complaint, AI-interview new hires were rated 4.6/7, against 5.2/7 for workshop new hires (a 13% gap) and 4.4/7 at head-office sites.
We followed 154 teams in nine organizations for 24 months, classifying each by the score sheets of its hiring panels in the first twelve months, during which the teams hired 611 people. "Values-floor" panels (n=52 teams, 208 hires) screened for a short list of conduct standards and then scored a candidate's difference in background or working approach from the existing team as a plus. "Similarity-fit" panels (n=55, 214 hires) scored how closely a candidate resembled the team, and "unstructured" panels (n=47, 189 hires) used no written rubric. Over 24 months, values-floor teams formally changed 3.1 long-standing practices per team after those practices were questioned, against 2.4 for similarity-fit teams (a 29% gap) and 2.5 for unstructured teams (a 24% gap). Complaint categories that recurred for three or more quarters made up 18% of values-floor teams' complaint categories, against 23% (a 28% gap) and 22% (a 22% gap). Conduct cases ran at 1.7 per 100 employees a year in values-floor teams, 1.6 in similarity-fit teams, and 2.1 in unstructured teams. Values-floor teams logged more disagreements escalated to the team lead in the first six months (4.6 per team versus 3.7, a 24% gap), and their new hires left more often in the first year (14% versus 11%). Among values-floor teams that paired each new hire with a peer for six months, first-year departures were 13%, against 16% in values-floor teams that did not.
We followed 487 newly hired employees in 64 teams across eight organizations through their first year. In "reviewed-log" teams, newcomers kept a short record of team practices that puzzled or surprised them during their first ninety days, and team leads went through it with them at days 60 and 90, with any item left open due a recorded response within four weeks (n=171). In "open-door" teams, newcomers were told they could raise anything at any time (n=168), and in "no-practice" teams, nothing was said either way (n=148). Reviewed-log newcomers recorded 7.4 items on average; 63% were answered with an explanation, 9% led to a change, 7% to a trial of an alternative, and 21% went unanswered. At six months, reviewed-log newcomers rated how far their perspective was valued at 5.3/7, against 4.5/7 for open-door newcomers (an 18% gap) and 4.3/7 for no-practice newcomers (a 23% gap). Among reviewed-log newcomers, this rating was similar whether none of their items had led to a change (5.2/7) or at least one had (5.4/7). In reviewed-log teams, every credited change arose from a logged item. Practice changes credited to a newcomer's observation within six months came to 0.67 per newcomer in reviewed-log teams, 0.54 in open-door teams (a 24% gap), and 0.52 in no-practice teams (a 29% gap). Team leads spent 3.4 hours per newcomer on the reviews, and at three months existing team members in reviewed-log teams felt more second-guessed (3.9/7, against 3.2/7 and 3.1/7), a difference gone by twelve months.