marklewin1's avatar
marklewin1
Valued Supporter
2 months ago

The Shine of AI Blinds Many Planners

The hype around artificial intelligence is louder than a footy crowd at State of Origin. Advisers see demos of Copilot writing recommendations in seconds and start thinking “problem solved.” Yet, just like a GPS can’t fix a car with no fuel, AI can’t fix a database that’s full of duplicates, half-filled fields and cryptic file names. At Back Office Hero (BOH) we live by a simple rule: systems first, software second. Until your data is clean, AI will only automate the chaos.

Data Hygiene: The Hidden Asset in Your Practice

Data hygiene isn’t glamorous, but it’s the bedrock of every profitable advice firm we benchmark. When client information is complete, current and consistently tagged, three good things follow:

  • Advisers trust the dashboards.
  • Support staff stop re-keying or second-guessing figures.
  • Compliance checks become click-through, not treasure hunts.

That foundation lets AI do the clever stuff - summarising three years of file notes, flagging lapsing insurance, or spotting a missing Fee Disclosure Statement - without spitting out junk.

What “Clean” Looks Like in a Financial Planning Firm

  • Each household has one master record - no “Smith Family” v “John & Mary Smith” duplicates.
  • Every phone number, e-mail and entity field is filled or deliberately blank (and noted as such).
  • File names follow a naming convention: 2025-05-13_Review_Brett Jones.pdf, not Doc123.pdf.
  • Rolling three-year history of goals and advice is recorded in plain language, not planner shorthand.
  • All data sits in one source of truth (Xplan, AdviserLogic, Worksorted), while SharePoint holds the supporting documents.

When those boxes are ticked, Copilot can mine the lot and deliver insights you can bank on.

ive Risks of Skipping the Cleaning Step

  • Bad recommendations – AI summarises an outdated SOA and you miss a change in objectives.
  • False compliance alerts – missing data fields trigger red flags that waste everyone’s time.
  • Reputational damage – clients see errors in meeting recaps and wonder what else you’ve missed.
  • Licensee headaches – sloppy records breach data governance rules and invite audits.
  • Money down the drain – you pay for AI tokens, consultants and add-ons that never reach ROI because the inputs are rubbish.

Simple Steps to Start Scrubbing Today

  1. Run a duplicate search across your CRM and merge double-ups.
  2. Agree a naming convention for folders and documents, then lock it in with a SharePoint template.
  3. Create mandatory fields for core client data - Copilot should never have to guess marital status or risk profile.
  4. Assign a Data Champion (one of our favourite BOH roles) who checks weekly reports and chases gaps.
  5. Adopt a “one touch” mindset - the person who learns the data makes the update, not the next person down the line.

Tackle one client segment per fortnight and you’ll be amazed how quickly the noise disappears.

AI Loves a Good System - And So Do Auditors

Once the scrub-down is done, Copilot can:

  • Generate a rolling “client story” timeline in 30 seconds.
  • Cross-check SOA, ROA and fee records against ASIC’s best-interests checklist.
  • Highlight dormant opportunities - unused pension cap, stale estate plan, or unmapped goals.

Just as importantly, every AI insight is traceable back to a clean data point, giving auditors confidence that your advice is bullet-proof. That’s the BOH mantra: process over heroics, evidence over memory.

Final Word: Diagnose Before You Prescribe

The allure of AI is real, yet meaningful results depend on disciplined data. Prioritise cleaning first and automation second; with a solid foundation, Copilot can serve as a reliable extra set of hands - without the risk of building on shaky ground.

Mark Lewin

Back Office Hero

May 2025

3 Replies

  • Great article marklewin1​, I couldn't agree more. It's going to be hard for many to take advantage of the AI / BI tech wave, unless they put in the effort now to fix data issues and implement a solid data entry protocol process. 

    Like you said: :"Until your data is clean, AI will only automate the chaos." 💯

  • luke.watson3's avatar
    luke.watson3
    Icon for Advisely Contributor rankAdvisely Contributor

    Couldn’t be more true…sh!t in = sh!t out.

    The source of truth removes confusion, and assists with decision making