Welcome back. This week the industry sharpened. Northern Trust's biennial survey of 300 asset management leaders, published on Tuesday, found product rationalisation has become a named strategy: the share of firms prioritising product reduction rose from 5% to 28% in two years, while outsourcing intentions more than doubled, from 18% to 39%. The vendors read the same memo: Anthropic built its first complete offering around the financial advisor's workflow, Broadridge pushed digital assets into US wealth management, and Automic launched a bundled fund platform with Northern Trust's custody behind it. The regulators sharpened too: the first deadline on the AIFM regime consultation passed on Friday, the SDR entity reports land on 2 December, and T+1 expectations are cascading into national rulebooks. None of it makes front pages. All of it redraws what your operating model owns. And underneath it all sits the question in this week's second insight: every firm now deploys AI, yet BCG's study of more than 1,250 firms finds only 5% extracting value at scale. Their answer is the 10-20-70 rule, and the 70 is the part nobody funds.

The industry's sharpening instinct, in one survey. Northern Trust's Driving Growth in Asset Management 2026, the biennial follow-up to its 2024 study, surveys 300 CEOs, CIOs, operations directors and similar leaders across North America, EMEA and APAC, and the shifts are structural. The share of firms planning to achieve their priorities by outsourcing non-core activities more than doubled, from 18% in 2024 to 39% in 2026. Offshoring is now the leading cost-control measure, cited by 69%, with outsourcing non-core activities at 42%. Product expansion as a priority fell from 60% to 47%, while product reduction rose from 5% to 28%. On AI, the finding is absolute: every respondent reported deploying AI in some form, led by data accuracy and quality control, document management and research. The two soft spots are the telling ones: nearly half name consolidating data from multiple sources as their biggest data challenge, and sourcing and aggregating investment analytics is the leading front-office headache. Ryan Burns, head of Asset Managers and Private Markets, Americas, frames the shift as growth "with greater discipline and a sharper focus on the areas where they can differentiate," extending successful strategies through "ETFs, semi-liquid funds and collective investment trusts." Angelo Calvitto, head of Asia Pacific, puts the operating model in one line: "Managers are doing more with fewer, more strategic partners."

Anthropic builds the advisor's operating layer. On Monday 14 September, Anthropic launched Claude for Financial Advisors, its first complete product built around a single financial-services role: meeting prep, portfolio analysis and compliance checks, connected at launch to Orion, Addepar, Envestnet, Charles Schwab, iCapital, Wealthbox and Zocks, and developed with Dynasty Financial, Schwab and Vanguard. The strategic read for operations teams is not the product. It is the pattern: the AI labs have stopped selling models and started selling the workflow around a role, which is exactly the layer asset managers used to assemble themselves. When the vendor ships the role's operating layer, the buyer's job shifts from building it to integrating and governing it.

Broadridge takes digital assets into US wealth. On 14 September Broadridge expanded its next-generation digital assets platform to US wealth management firms, with Anchorage Digital and Galaxy Digital as initial ecosystem partners. The pitch is deliberately unromantic: cryptocurrencies and tokenised securities sitting inside the existing advisor desktop, client statement and back-office control stack, with advisor-led and self-directed paths rather than a separate crypto-only account experience. Six days after it put institutional workflows on tokenised rails (issue 006), Broadridge is extending the same plumbing toward the books-and-records and communications workflows of wealth firms. The operating layer between traditional and digital assets now has a vendor racing to own it.

Something worth naming: the industry's growth strategy is being rewritten as an operating model strategy, and the survey is the evidence.

Product reduction is now a named strategy. For as long as the surveys have been running, "expand the product line" was the default answer to the growth question. Not any more: the priority fell from 60% to 47% in two years, and product reduction rose from 5% to 28%, a near six-fold increase. Managers are choosing fewer, better-resourced products over broad shelves, and extending winners into new wrappers, ETFs, semi-liquid funds, collective investment trusts, rather than launching new strategies.

Outsourcing intentions have doubled, and offshoring leads the cost playbook. 39% of firms plan to achieve their priorities by outsourcing non-core work, up from 18% in 2024; 69% cite offshoring as their top cost-control measure. Calvitto's framing is the strategy in one sentence: "The goal is not only lower cost, but a more straightforward operating model with stronger control, better quality and the scale to support future growth."

AI is now universal, which changes the question. Every respondent deploys AI in some form. Every one. The differentiator is no longer adoption; it is what the AI sits on. Nick Gilbert, head of Asset Servicing, EMEA, says it plainly: AI's "value will depend on the quality, governance and accessibility of the data beneath it. This is not just a technology issue; it is an operating model and resilience issue." Nearly half the market says consolidating data from multiple sources is its biggest data challenge. Universal AI on fragmented data is how you get confident errors at scale. Now set that against BCG's September 2025 study, The Widening AI Value Gap: of more than 1,250 firms across 68 countries, only 5% are achieving AI value at scale, and 60% are achieving no material value at all despite substantial investment. Universal adoption, almost no value, and the reason is the ratio BCG themselves lay down: 10% algorithms, 20% technology, 70% people and process. The 70 is the part that never gets a line item, and it is the subject of the second insight below.

Architecture Canvas lens: sharpening is an operating model act, not a strategy act. Every product you retire drags processes, data feeds, tax wrappers and obligations behind it. Every process you outsource moves complexity to the boundary. Every agent you deploy reads the data layer you have, not the one you wish you had. The firms that treat sharpening as three disconnected projects, a product cull here, an outsourcing deal there, an AI pilot over there, will find the complexity did not leave. It relocated. More on that in the Canvas Corner.

Automic bundles fund administration and custody for the mid-market. On 14 September Automic Group launched its enhanced fund administration and custody platform in Sydney, supported by Northern Trust's global custody network, spanning custody, trade processing, corporate actions, investment accounting, tax and investor administration on a single data layer. It already supports more than 120 fund managers and 850 funds, roughly A$100 billion under administration. CEO Mal McHutchison's framing is the market's: a "complete and connected operating model" that "eliminates the traditional fragmentation across registry, administration and custody services." The Australian story matters beyond Australia. It is the template for mid-market managers everywhere: an administrator-bundled operating model with a global custodian underneath, sold as a product rather than assembled by the client.

Fund+ sells regulated fund infrastructure as a platform. On 10 September Fund+ launched in Abu Dhabi Global Market, an FSRA-regulated platform that lets managers, sponsors, wealth platforms and family offices launch, operate and administer vehicles without building the compliance, operations, documentation and technology stack from scratch. Read it next to Automic and the pattern is the same: the pre-packaged operating model, regulator included, is now a product category.

The private markets CIT supply chain assembles. On 14 September Great Gray Trust Company selected iCapital as its preferred partner for private markets manager evaluation, due diligence and portfolio construction, to develop asset-class-specific collective investment trusts across private equity, private credit and private real assets for defined contribution plans. Great Gray keeps the fiduciary, trustee and servicing spine; iCapital brings manager selection, allocation design and investment lifecycle management. The operational significance: the hardest parts of private markets, capital calls, liquidity management, valuation cycles, are being assembled into retirement-grade plumbing. Every layer that becomes standardised is a layer the buyer no longer builds, and a capability the assembler owns.

The AIFM reform consultation's first deadline passed on Friday; the main one is five weeks out. CP26/28, the FCA's rebuild of the UK AIFM regime, closes in two tranches: responses to the discussion chapters, covering depositary, prime broker and business restriction proposals, were due Friday 18 September, and the main consultation, together with the prudential discussion chapter, closes Thursday 22 October after an extension. Implementation of the new regime is envisaged for 2028. For operations teams, the deadline that matters is not the compliance response. It is what the new regime will ask firms to evidence about how they actually operate, and whether your operating model can produce that evidence without a consultancy engagement. The firms that have processes mapped, owned and re-certified will answer in a week. Everyone else will discover how much of their model lives in people's heads.

The SDR entity report lands on Wednesday 2 December. Under the FCA's SDR regime, asset managers with more than £5bn in AUM must publish their first entity-level sustainability report by Wednesday 2 December 2026; PwC's summary of the phased dates is the clearest reference, with the £50bn-plus tier having published in December 2025. For operations, this is a data-assembly and disclosure-controls exercise wearing a sustainability costume: entity-level disclosures span the whole business, not a fund range, which means data chains, definitions and sign-offs many firms have never had to produce at firm level. Ten weeks is not long.

T+1 expectations cascade into national rulebooks. On Wednesday 26 August, CySEC issued Circular C797, relaying ESMA's T+1 public statement to regulated entities: the first regulatory milestone remains Monday 7 December 2026 for allocations and confirmations, and firms are expected to assess dependencies across the trading and settlement chain and allocate resources now. Cyprus is one regulator; the pattern is general. ESMA's deadlines are being transposed by national competent authorities one circular at a time, and the firms treating Monday 11 October 2027 as the only date will meet December's requirements the same week they discover them.

Moves from the past two months.

Two operating model insights this week: The Sharpening Clause, and The 70.

The Sharpening Clause. Every operating model carries a clause nobody writes down: the model is only as simple as its owners want it to be, and simplification is not subtraction. When a firm sharpens, cuts products, outsources processes, deploys AI, it does not remove complexity. It relocates it. The clause says complexity always has an address, and you have to know it.

This week's survey describes an industry mid-sharpening. Product reduction is a named priority for 28% of firms. Outsourcing intentions have doubled. Every firm uses AI. Each of those moves is rational on its own. The clause is what happens at the boundaries. Cut a fund and its tail, the processes, data feeds, tax wrappers, regulatory filings, does not disappear; it lands on the surviving funds and the teams that support them. Outsource a process and its complexity moves to the provider boundary, where oversight is precisely where firms are weakest: in the FCA's financial crime review, only 36% of firms outsourcing CDD could evidence full oversight of the provider's onboarding. Deploy AI across a fragmented data estate and the complexity migrates into the model's answers, where it is hardest to see.

Three things to do:

The firms that sharpen with the clause in mind will find the market rewarding simplicity with scale. The firms that sharpen without it will discover, a year from now, that nothing was subtracted. It was just moved somewhere they do not look.

This week's companion piece on opmodal's Insights page goes deeper: Simplification is not subtraction: where complexity goes when the operating model sharpens.

The 70. The second insight starts from BCG's 10-20-70 rule: 10% of the value sits in the algorithms, 20% in the technology, 70% in the people and the process. The 5% of firms extracting value at scale have funded that ratio. The 60% extracting almost nothing have inverted it without noticing.

Why does the 70 go unfunded? Because it has no line item. A licence lands on a purchase order. "Two approval gates removed after a fight" fits nowhere in a budget template, so it becomes "change management" and quietly slips. Underneath sits a harder problem: you cannot price what was never structured. Nobody can scope the redesign of a workflow that has never been mapped, so the 70 gets waved at in the business case and dropped the moment the numbers are built. The ratio is not disbelieved. It is starved.

And the mapping is genuinely absent. The documented process and the way work actually happens are two different things. The real process is distributed across the people doing it: the workarounds, the exceptions, the duplicate steps and informal handoffs that never reach the process map. In a Lloyd's managing agent that gap is the operating model itself: hundreds of Visio swimlanes and Word procedures, PDF'd, unsearchable, owned by nobody, and out of date the day the last change programme closed. Before "where can we apply AI?" comes a prior question: do we actually know how this work gets done today? Automating a poorly understood process does not remove the problem. It makes the problem faster.

The fix is to give the 70 an address: turn the documentation you already have into one connected operating model, so that workflow redesign becomes scoped work: this process, these steps, these gates, this downstream impact. At that point the 70 is a cost code, not a hope. If your AI budget is 80% technology, you have funded the easy third and starved the part that pays.

The second companion piece goes deeper: Budget for the 70: why the largest share of AI value has no line item.

Target State is published weekly by opmodal. Subscribe free at target-state.beehiiv.com. Look out for the paid tier opening this autumn.

Architecture Canvas is a central component of the methodology developed by opmodal for capturing, governing, and improving operating models in complex organisations.