 |
I read an interesting article in last weeks Guardian about AI targeting of that girls school in Iran and possibly how that went pear-shaped. While most of what papa Trump touches goes wrong this is a surely data shaped cock-up.
The bit in that article lands harder than all the usual breathless AI chat, because it gets to the real problem. It was not some magic machine turning evil on its own. Surprise, surprise it was stale data, bad assumptions, poor governance and human beings building speed on top of shaky foundations. That is the bit people in SocialHousing and Asset Management ought to be paying attention to, because we are in danger of doing our own smaller-scale version of exactly the same thing.
No, social housing is not a war zone (I mean now and again it may feel so!), and I am not being melodramatic. But the underlying lesson is the same. When old, wrong, shanky or incomplete data is fed into newer, faster, more sophisticated reporting tools, the output can look clever while being fundamentally unsound. The system appears confident. The dashboard looks polished. The report has plenty of decimal places. The presentation to the board feels reassuringly modern. Meanwhile the underlying asset data might be a mess. Only our asset management colleagues at the sharp end may actually realise this.
That is our Achilles heel.
I see far too much discussion at the moment about clever tools, predictive analytics, AI, optimisation engines, automated investment planning and all the rest of it, with nowhere near enough honesty about the quality of the data sitting underneath. If your component records are patchy, if installation dates were guessed fifteen years ago during a rushed LSVT stock condition exercise, if replacement history, or subsequent SCS's have not been consistently updated, if attributes have been bulk-loaded from old systems with half the meaning stripped out, then wrapping that in shiny new software does not create truth. It just creates faster nonsense.
And housing does love a bit of faster nonsense at times. We have all seen it. A landlord buys a new asset system, or switches on a new reporting layer, and within five minutes someone wants a thirty-year investment profile, a full life cycle costing model, and a heat map of future liabilities by estate, archetype and component. Lovely stuff on paper. But if your kitchen data is wrong, your roof data is vague, your windows are classified three different ways (by spot cost, by unit, by archetype, or a bit of all?), and half your components are defaulted to generic dates because nobody knew the real answer, then that LCC report is not a strategic asset. It is an expensive comfort blanket.
That is the danger with life cycle costing in particular. People often talk about it as if it is some kind of grown-up, evidence-led answer to long-term planning. In principle, it can be. In practice, it is only as good as the component data, assumptions and replacement rules beneath it. If the dates are wrong, the quantities are wrong, the useful lives are generic, or the survey data is incomplete, then the report can give a completely false sense of control. It can tell you with great confidence that spend will peak in year eight, flatten in year twelve and rise again in year nineteen, when in reality it is built on a pile of half-remembered survey notes, inherited coding structures and wishful thinking.
This is where people get seduced by sophistication. They think the answer to poor information is a better tool. Sometimes the answer is a torch, a clipboard, a proper data model and a grim few months of sorting out the basics. That is less exciting, of course. Nobody gets invited to a conference to say, “we cleaned up our component hierarchy, tightened our validation rules and stopped pretending unknown dates were known.” But that dull work is often worth more than the glossy AI strategy that comes afterwards.
In the article I highlight, failure was not simply “technology”. The failure was letting a system run at speed while relying on information that no longer matched reality. That should ring bells in housing. We do it when asbestos records are incomplete. We do it when stock condition data is old. We do it when compliance statuses depend on weak integrations. We do it when repairs history does not feed asset intelligence properly. We do it when planned works teams, compliance teams and housing management all hold slightly different versions of the truth and nobody is quite sure which one drives the report to the board.
Then along comes a new supplier or a new platform promising insight. Suddenly old problems are dressed up in modern clothes. Instead of one vaguely unreliable Excel spreadsheet, you get a dynamic dashboard. Instead of a crude five-year plan, you get scenario modelling. Instead of a rough estimate, you get “predictive” forecasts. But prediction based on bad source data is not insight. It is theatre.
That is the bit I think the sector still does not say loudly enough. Data quality is not some boring back-office technical issue. It is operational risk. It affects compliance, investment, tenant safety, financial planning and organisational credibility. If you do not know what components you have, where they are, when they were installed, what condition they are in, and whether that information has been properly maintained over time, then the fancy reporting layer is not your salvation. It is just a better quality mirror for your confusion.
And to be fair, this is not always laziness or incompetence. A lot of landlords are carrying legacy data structures and systems that were never designed for today’s expectations. They have been through mergers, outsourcing deals, stock transfers, system migrations, rushed mobilisation exercises and years of under-investment in proper data stewardship. Bits of the truth are scattered everywhere. The asset system says one thing, repairs says another, finance codes something differently, and the survey programme has never fully caught up. Then somebody asks for a fully costed decarbonisation pathway or a stock reinvestment plan by component and tenure, and wonders why the answers wobble.Because the data wobbles, so them we assume the system is wibble.
So before anybody gets too excited about AI, digital twins, predictive replacement models or miracle dashboards, there is a simpler question worth asking. Do we trust the component data? Not “is there some data”, not “can the system produce a report”, but do we genuinely trust it enough to make expensive, high-impact decisions from it? If the answer is no, or even “sort of”, then that is where the effort belongs.
There is no glamour in saying your first priority is data assurance, better survey discipline, tighter updates from planned works, cleaner integrations and proper governance around defaults, assumptions and exceptions.
But that is the grown-up answer. In housing, as in every other serious field, speed built on stale data is dangerous. The consequences may differ, thankfully, but the principle does not.
We should remember that every time someone waves a shiny new tool in front of us. Clever software cannot rescue weak asset knowledge. It can only amplify it. And if we are honest, that is probably the real lesson here. Our problem is not that housing lacks sophisticated systems. It is that we still too often ask sophisticated systems to compensate for basic uncertainty. That never ends well
Related Post: As a client told me recently, the right Critical Friend, will potentially save you more money than you spend.
I would be pleased to connect with you on LinkedIn - http://uk.linkedin.com/in/tonysmiththathousingitguy Message
me with any issues or queries, you would like to be explored in this
blog. We generally receive a couple of suggestions each month.
Mansun - Dark Mavis
(c) Tony Smith, Acutance Consulting www.acutanceconsulting.co.uk 07854-655009
Access a quick list of our Social Housing ICT blog posts here , what projects do we get involved in? Access them here
Could we help you or your organisation? Our contact details are here , get in touch we will be pleased to chat about your problems and help with your organisation issues.
File
Under:
#HomesForBritain,#HousingDAY,#InternetOfTenants,#Shout,#UKHousing,
07854655009, 1st Touch,1stTouch,360,365Agile,3squared,4Js,
ACL,ALMO,AMS,AX,Aareon,Abritas,Academy,Accuserv,Accuserve,Active
Housing,ActiveH,ActiveF,ActiveFinance,ActiveHousing,Acutence,Advanced,Affinity,Aidenn,Agile,Agile365,AHB,AirWatch,Alscient,Alfresco,Alignment,Allpay,Amazon
Web Services,Anite,Apex,App,ArchHouse,Archouse,Asprey e-state pro,Asset
Management,Associates,Aurora,Average IT Costs,AWI SX Integration
Toolkit,AWS,
BI,BO,BPR,BancTec,BigChange,Blockwise,BluTek,Bluebox,Blueprint,BrikHousing,Browser
Applications,BC, Business Central, Business Objects,Business
Planning,Business Process Review,Business social networking, CACI,Capita
One Housing,CBL,CCS IT Keystone
CCSIT,CEDRM,CHICS,CHR,CIH,CMS,COA,CORE,CPL,CRM,CRM2013,CRS,CTI,CTX,Cadcorp,Capita,CapitaOne,Capita
One,Capital Management,Cashflow,Castle,Castleton,Castleton
Technology,Cedar Open Accounts,Cerrus,Change,Charges,Chartered Institute
Of Housing,Cheaper Housing IT,Chics, Citrix,Civica,Civica CTX,Civica
Cx,Civica Genero,Civica Saffron,Clearview,Clik,Cx Assets, Cloud
Dialogs,CloudDialogs,Coactiva,Codeman,Comino,Commontime,Community Reward
Services,Company,Competitive Dialogue process,Component
Accounting,Consilium,Consolidation,
Consultancy,Consultant,Consultants,Contact Manager,Context,Contractor
Systems,CorVu,Cost Reductions,Covalent,Crystal Reports,Customer
Relationship Management,Cx,CxFeedback, DRS,Deeplake,Datatank,Designer
Software,Development
Systems,Director,Document,Documotive,Docuware,Dynamic
AI,DynamicAI,Dynamics 365,D365,Dynamics365,Dynaway
ECMK,EDM,EDRMS,ERP,ESRI,Elmhurst,Enghouse
Interactive,England,English,EnterpriseBI,Estatecraft,Esuasive,Esynergy,Etive,Exhibition,Exponential-e,
F&O, Facebook,Factorwise,Fast Lean Smart, Field Service
management,Finance,Financial
Systems,Financials,FLS,Footprint,Forms,Freezes,Fusion,Fuzzlab Gas
Tag,G-Cloud,GCloud,GDPR,GGP,GIS,GasTag,Genero,GeoSolveIT,Getting best
from,GoTonySmith,Grasp,Grip,Group Apex, HCL,HFI (Housing Financials
interface),HG,HRA,Hardware,Hitachi
Systems,Hitex,HomeMaster,Hometeam,HouSys,House,Housemark,Housemark
survey,Housing,Housing Contact,Housing Group,Housing Insight,Housing
Management,Housing Management Consultant,Housing Partners,Housing
Portfolio Management System,Hub Asset Management,Homeswapper,Housing
jigsaw,HousingOne,Housing-One, Housing Support
Pro,HousingIT,HousingSupportPro,HyperOptic, IMS,IPC,iProperty
Cloud,IT,IT Budget,IT Training,ITIL,Impact Response,In,In4,In4Systems
Promaster,InfinityGroup,InfoBoss,Infoshare+,InHouse,InMotion,InMotion2015,InMotion2016,InfoBoss,Infoflow,Information,Information
Technology,Information@Work components,Informix,Innovation,
Informetis,Inphase,Inside Housing,Insight,Internet
Portal,Internetalia,Invu,Ireland,Irish,itLab,InclineIT,
Keyfax,Keylogic,Keypera,Keystone,Kirona,Kypera,Keynamics,
Landlord,Ledgers,Linkedin,Liquid,Locality,Localz,Looking Local,
M3,MadeTech,Manifest Software
Solutions,MAVIS,MD,MDM,MI,MIS,MIS-AMS,MISCS,MRI,MS Dynamics,MRI,MS
Dynamics CRM2011,Microsoft365,microsoft 365,MWL,Management,Management
Server,Manifest,Measuring,Mebus,More IQ,MoreIQ,Microsoft Dynamics
365,Microsoft dynamics GP 2013,Miracle,MobileIron,Mobysoft,Monopoly
board images and pictures,Montal,Mr Void,MriEngage,MrVoid,MoreIQ,
MRI,MRI/Capita,MRI Housing One, NDL,NINTEX,NINTEX workflow,NPS,NPS
ASSIST,NPS
Housing,NPS Job Manager Mobile,NROSH,Natural,Neighbourhoods and
Communities,Nintexworkflow,Northgate,Northgate Codeman,Northgate Public
Services,Notice, OA,OGC Buying Solutions,OJEU
Limits,ORS,Ohms,OmFax,Omni,Omniledger,One,OneAdvanced,OneServe,Open
source,OpenContractor,OpenHousing,Opti-time,Options,Optitime,Oracle,Orchard,Outsource
and outsourcing,OutSystems, PIMMS,PIMSS
Data,PM,PRINCE2,Paloma,Pamwin,PanConnect,People
Value,Peoplevalue,PfH,Pick,PlanForm,Planned
maintenance,Plentific,Plus,Portfolio Management
System,PowerObjects,Pro,Pro-Points,Prodo,Progress,Promaster,Propeller,Propoints,Proval,Providers,Pyramid,
QL,QLX, QLF, QL Yuneo,QuantSpark,Qube,QLX,Qlikview,QueryView,Queryview
reporting,Quiss, RM865,ROCC,ROCC
Orchestrator,RPs,RSL,Rave,ReAct,Reality,RedkiteCRM,ROCC Attend,Red
Olive,Registered,Registered Social,Rent
Increase,RentSense,RentSenseLite,Repairfinder,Reporting,Reports,RobotAutomation,ROCC,
Rocket,Rubixx,
Salesforce,Salesforce.Org,SAP,SASSHA,Sava,SDM,SHBVN,SM,SOTI,SP,SQL,SQL
Open
Housing,SQL Reporting,SQL
Server,Saffron,Safron,Saturn,Scenario,Scotland,Scots,Scottish,Scout,Sequoa,Serengeti,Serros,Server,Service,Services,Servitor,Sharepoint,Sim,SimPro,
Simdel,Simdell,Slash and Burn,Social Housing ICT,Social Housing
Software Applications,Social Media,Software,Software solutions,
Solvares, Spotlight
Service,Spotlightservice,Star rating,Stores and
Stock,Strategic,SunAccounts,Sunguard,Surveys,Streetwise,
Swordfish,Sx3,Symatrix Human,System alignment,Systems,Systemwise,
T-Files,TEAMS,TED,TFiles,TP Tracker,TSG,Tagish
Casework,Tagra,Task,TAIM,Techlabs,Technologies (India) PVT
Limited,Telecetera,Template,Terminal Services,That,Three
Star,Tilt,TiltAffinity,Today,Tony Smith,Tony Smith That Housing IT
Guy,TonySmith that housing IT,TonySmithHou,TonySmithHousing,
,TonySmithHousingITguy,Total,Total
Mobile,Totalmobile,Trace,Tribal,Twitter, U2,UC,UK,UK
Housing,Ukhousing,UniClass,Unidata,United
Kingdom,Universal,Universalcredit,Universe,Unrest, Valueworks,Van
Stock,Vantage,Vantage
Sentinel,Version,Virtualisation,Visitour,Visualmetrics,Voice,Voice and
data,Voicescape,Voluntas, Wales,Welsh,Wheatley,Windows Server,Workflow
and,Work Hub,Works Connect, XML,Xen App,XenApp,Xmbrace, Yuneo,TS
Acutance Consulting, ...
