Purposeful Configuration: Why the Firms Winning in Q2 Are Running Finished Tools, Not Better Ones
Market Intelligence Analysis for Professional Services Leaders
Week of May 25, 2026
1. The Earnings Signal You Should Be Watching
The most useful data point this week didn’t come from a government report or an index. It came from two earnings calls that painted opposite pictures of the same market.
Adecco Group posted +5.3% organic revenue growth in Q1 2026. Americas came in at +15% organic — the strongest regional result in the quarter. Their CEO pointed directly to why: agentic AI deployments on their digital platform are improving fill rates and time-to-fill in new markets. Flexible placement revenue grew 6%.
GEE Group posted Q2 2026 revenue of $19.5 million — down 20.5% year over year. Contract staffing revenue fell 24.2%. Their CEO named AI directly as a contributing factor, describing a “dampening effect” on hiring plans at many client organizations.
Same macro environment. Same quarter. One firm up 15% in Americas. One down 20.5% overall.
This split is worth sitting with, because it’s not about market access or firm size. Adecco built AI deployment into its delivery operations — purposefully, at scale, with measurable outcomes. GEE Group is absorbing the uncertainty its clients feel about what AI means for their headcount. That’s not a technology gap. It’s a configuration and positioning gap.
And it’s showing up in earnings now. Not as a forecast — as a result.
This matters because the ASA Staffing Index hit 88 this week, with staffing hours reaching a year-to-date high. The demand signal is real. The question is whether your operating infrastructure is built to capture it. In this market, having the tools and running the tools are not the same thing.
2. What Purposeful Configuration Actually Means
Newbury Partners used a phrase in their May 20 post that deserves wider circulation: purposeful configuration. They applied it to BI portals, but it describes a problem running through every layer of the staffing tech stack.
Here’s the simplest version: a BI portal set up for reporting shows you what happened. A BI portal configured for decisions shows you what to do about it. The technology is often identical. The setup is not.
Think about it this way. If your dashboard shows call volume and submission counts but not conversion rate by rep or pipeline velocity by stage, you have a reporting tool. It confirms effort. It doesn’t drive decisions. The configuration didn’t finish.
Newbury ran this argument through four layers this month. Their May 13 post on BI risk signals: dashboards surfacing activity metrics without conversion or velocity data. Their May 15 CRM audit: a four-question diagnostic that exposes the silent configuration debt sitting in most HubSpot installs — pipeline stages that match the default rollout, not your actual sales motion. Their May 8 piece on ATS implementation recovery: systems that completed installation without finishing the workflow design that saves recruiter time.
Four layers. Same gap at each. The tool is on. The configuration isn’t done.
In a market where industrial staffing hours are running +9% year over year, that gap has a direct cost. Seats stay unfilled longer than they should. Candidates drop off between offer and start. Pipeline reviews become volume conversations instead of conversion conversations.
Randstad Digital’s 2026 workforce research names a related problem on the people side — they call it the Productivity Paradox. Companies are deploying AI platforms faster than they’re building the capability to use them. One in four tech professionals surveyed across 35 markets has already quit a job because their employer failed to provide structured upskilling. 52% are seeking training independently because internal programs can’t keep pace.
Configuration isn’t just about your tech stack. It runs through the people operating it. Both need to be finished before the market rewards precision over broad spending.
3. The AI Governance Gap Is Wider Than You Think
Littler’s 2026 annual survey of approximately 300 C-suite leaders produced a number worth marking: 84% expect AI-related policy or regulatory changes in the next 12 months. That’sdouble the share who said the same last year. AI governance has officially overtaken both immigration and DEI as the top area where employers expect regulatory action.
The good news: 68% of organizations now have a formal AI policy — nearly double last year’s rate. The harder part: just over half of those firms have actual restrictions on what information can be entered into AI tools. The gap between “we have a policy” and “the policy covers the real risks” is where legal exposure is accumulating right now. Most firms don’tknow which side of that gap they’re on.
Littler’s 84% number reads differently after a week at the World Employment Conference in Toronto. Manpower Group CEO Jonas Prising opened the event with a 72% global talent shortage figure drawn from 100+ billion data points across 80 countries — and paired it with an explicit call for ethical AI guidance covering explainability and reproducibility of candidate ranking. The Global Worker Classification Forum walked through the EU Platform Work Directive, where algorithmic management is now a legal employment indicator, Ontario’s Digital Platform Workers Act — already binding — and ILO Geneva negotiations starting next month on an international platform-work standard. The takeaway: the 84% C-suite expectation of regulatory change isn’t a forecast. It’s the trailing edge of infrastructure that’s already operating in other markets. Firms building first-version policies today are catching up to where international competitors were a year ago.
For staffing firms, this plays out at two levels.
Your own operations first. If you’re using AI in sourcing, candidate communication, or screening — and most firms are, in some form — you need documented input restrictions, bias monitoring protocols, and adverse action procedures. Not because enforcement has arrived. Because 84% of the C-suite leaders you work with expect it to arrive within the year.
Then your client relationships. Your clients are running the same governance gap audit on their own organizations right now. A staffing firm that shows up with a formal AI governance framework for its delivery operations will have a very different conversation than one that doesn’t. That’s not just compliance. It’s a trust signal — and in a market where trust drives contract extension and referral volume, trust signals compound.
The firms that close this gap in Q2 won’t be scrambling to retrofit policies when the first enforcement action lands. They’ll be the firms clients call when they need to understand what compliant, AI-assisted hiring looks like in practice.
4. The Talent Workflow Bottleneck
Industrial hours are up 9% year over year. Commercial hours are up 7%. These are the highest-volume segments right now — and both are climbing while office/clerical continues to contract at -4%.
Vetty’s May content series ran a consistent thread: candidate drop-off, talent leaks, and compliance gaps in screening are downstream symptoms of a configuration problem, not a sourcing problem. When ghost rates rise in a hot industrial market, firms look upstream — job boards, sourcing channels, recruiter activity. The problem is almost always downstream: screening delays, unclear onboarding steps, or broken communication workflows between offer and start.
The data is consistent. Slow background check turnaround correlates with higher ghost rates. Fast, automated, mobile-first screening workflows correlate with lower ones. The fix isn’ta new sourcing channel. It’s a configuration audit of the steps between “offer accepted” and “day one.”
There’s a broader labor dynamic worth understanding here. Express Employment and Harris Poll found that 60% of hiring managers now say it’s more efficient to use AI for entry-level tasks than to hire and train a candidate. SIA’s recent research describes a shift from “how many people do we need” to “how should work be allocated across employees, contingent labor, and digital workers.” That reallocation is not happening uniformly across sectors.
Industrial and skilled trade roles are not compressing. Administrative and entry-level clerical roles are. If your placements sit on the industrial side, Q3 demand is already showing in the leading indicators. The seasonal screening workflow you build in the next two weeks is the one you’ll run at volume in July and August.
Signature Back Office’s post on building contract hiring pipelines before demand peaks makes the companion point: infrastructure built before you need it holds under pressure. Infrastructure built under pressure rarely does.
5. The Execution Discipline Problem
Butler Street published three posts this month that are worth reading as a series.
“Don’t Get Fooled By Fancy Tools” starts with the pipeline problem. Staffing teams invest in better CRMs, AI platforms, and dashboards without changing the prospecting, discovery, and follow-up discipline that determines whether pipeline is predictable. The argument is blunt: inconsistent fundamentals produce inconsistent pipelines regardless of how good the instrumentation is. Better dashboards make the problem more visible. They don’t fix it.
“Why Access to Information Alone Doesn’t Lead to Better Performance” goes a layer deeper. In a world where AI delivers answers in seconds, performance isn’t improving at the same rate. The gap isn’t information access. It’s knowing what to apply, when to apply it, and how to reinforce the behaviors that actually move results. Tools and training only create impact when paired with coaching and consistent execution reinforcement.
“Good Is Not a Permanent Place” makes the leadership case. The teams pulling ahead in 2026 are using smarter preparation, more disciplined QBRs, and daily accountability — not out-running a competitor with broader resources.
The through-line across all three: configuration runs through people, not just platforms. A CRM with finished configuration and a team still running inconsistent discovery discipline will still produce an unpredictable pipeline. Both layers need to be addressed. A new tool running on an old execution problem isn’t purposeful configuration. It’s a more expensive version of the same problem.
6. The Q2 Configuration Roadmap
The data this week points toward a clear 30-day sequence. It has three steps.
Run the diagnostic first. Answer Newbury’s four CRM audit questions honestly — pipeline stage alignment, workflow reliability, reporting output, data hygiene. Add the conversion delta column to your pipeline dashboard. Pull background check turnaround time against your ghost rate in your own data. Document every gap. Don’t start with the fix. Start with an honest read of where the debt is sitting.
Build before you need it. The Q3 seasonal screening workflow needs to exist before Q3 demand arrives. The AI governance policy needs to be documented before the next regulatory guidance drops. The industrial workflow optimization needs to happen while there’s still time to test it — not under July volume pressure. The firms that built infrastructure ahead of Q2 demand are the ones showing up at +9% year over year in industrial right now. The firms that waited are running manual processes in a market that doesn’t reward them.
Measure the right things. Conversion delta by rep. Placement rate by source. Retention by source. Background check turnaround versus ghost rate. The firms that built measurement infrastructure in Q1 and Q2 will make better decisions in Q3 — not because they have more data, but because their systems surface what matters rather than what’seasiest to count. That’s the difference between a reporting tool and a decision tool.
The Adecco/GEE Group split doesn’t close on its own. The firms absorbing client AI uncertainty will keep losing contract volume until they build delivery infrastructure that makes purposeful AI deployment visible to buyers. That’s a Q2 decision with H2 consequences.
7. What the Data Points Toward Next
June 5 gives us the May Employment Situation from BLS. Given the current trajectory — staffing hours at a YTD high, industrial at +9%, the Conference Board Employment Trends Index at 105.77 with temp help positive for four consecutive months, and initial claims near a historic low at 203,300 — the probability of another divergent reading is high. ASA elevated, BLS softer. Cross-reference it when it drops. ASA is the staffing-specific signal. BLS is context. Firms calibrating hiring plans against BLS alone are reading a different market than their clients are buying in.
The AI governance window is the one to watch most closely. 84% of C-suite leaders expecting regulatory changes within 12 months is a leading indicator, not a lagging one. The firms building formal policies now — input restrictions, bias monitoring, adverse action documentation — will have the compliance infrastructure in place when guidance arrives. The firms waiting will build it under deadline pressure, which is when compliance programs develop gaps.
Longer term: SIA’s research on workforce model redesign — the shift from headcount planning to work allocation across employees, contingent labor, and digital workers — is worth tracking as a Q3 and H2 signal. As clients redesign workflows around AI agents, the shape of contingent demand changes. Agencies that understand how work is being reallocated, not just how headcount is shifting, will be positioned for the roles the hybrid model can’t automate. That requires a different conversation with clients — and it starts with your own configuration being finished enough to credibly have it.
The demand is there. The configuration is the work. The window to do that work — before Q3 volume closes it — is now.
Based on verified intelligence from ASA, BLS, SIA/Bullhorn, Littler, Adecco, GEE Group, Randstad Digital, Express Employment/Harris Poll, Conference Board, Newbury Partners, Butler Street, Vetty, Signature Back Office, and World Employment Conference 2026. All data points current as of May 25, 2026. For detailed sourcing, see The Current weekly intelligence report
