Hi everyone!
I am currently in the process of developing a software program that pulls data directly from a client's bank account, automatically codes it and allows the user to report on the data. The software will be very secure and the data is provided by a business that is regulated by the SEC in America and follows all the normal regulations here in the UK such as PCI standards. The user simply enters their online banking details once and the data is pulled over every day (certain accounts require the user the enter an online banking security pin on each update). The idea is to create a tracking tool for those clients who will never maintain an Accounting system or who will never pay their Accountant to maintain one for them.
I would really appreciate some thoughts on this from Accountants in industry as I am currently at the stage where I need to invest significant funds!
Thanks in advance!
Replies (20)
Please login or register to join the discussion.
Automatic coding
Is it your intention that the coded bank deposits & withdrawals should form the basis of accounts & tax returns without the need for an external accountant?
If so I see a danger because the nature of an expense will depend upon the nature of the underlying business. What may be drawings in one business may be purchases for resale or legitimate business expenditure in another and of course capital expenditure may give rise to tax allowances (or not).
David
Machines aren't that smart
I'm wondering whether it wouldn't cost me time.
Machines aren't smart enough to deal with something out of the ordinary.
Established players
I think you'll find this a market which is already catered for, either with cloud based software which already allows bank data to be imported and quickly coded - Xero, ClearBooks etc.
Bankstream are already doing what you are developing.
So you'll have stiff competition.
Is the competition something you have considered?
Plus the process you will need to go through to get bank approval and possibly approval by accountancy bodies to give your product credibility.
third party intermediary
Bankstream is a competitor but they have only got three banks on board, we would have over 60. We wont need bank approval as we will be getting the data from a third party intermediary.
I expect I am very naïve, but which third party intermediary would have (for example) my bank data?
I expect he is referring to a third party such as Yodlee, who provide software that knows how to log in to various banks' systems. You give them your account details, including your 'memorable information' and their software will log in as you once a day and fetch any new transactions.
The reason why they don't need to be in partnership with the banks per se is that the software is emulating a regular user using a browser, clicking buttons, filling out forms etc.
There are of course major trust and security issues here.
To answer your question, no 3rd parties should have your data unless you have given them your full login details and explicit permission to fetch your records on your behalf.
.
Bankstream is a competitor but they have only got three banks on board, we would have over 60. We wont need bank approval as we will be getting the data from a third party intermediary.
I expect I am very naïve, but which third party intermediary would have (for example) my bank data?
I'll carry on waiting for OP's reply to my question.
Features lacking with existing suppliers (eg. Xero)
So far as I am aware none of the existing suppliers of Xero-like systems go beyond what are in effect bookkeeping and management accounts to produce the following really useful "extra mile":
Statutory accounts for UK Ltd Co'siXBRL accounts for Corporation Tax purposes
I suspect that people like Xero either have the above on a "wish list" or see no money in it (ie. development costs outweigh what they see as the potential extra income).
I am aware that Xero can already produce New Zealand statutory accounts, but what I need is UK statutory accounts plus iXBRL, then I'd look forward to being able to say goodbye to substandard and overly complicated accounts products such as Sage Accounts Production Advanced (SAPA for short, pronounced as "TRASHA" [or worse] by some of those who are new to it).
For simple accounts ...
... FreeAgent links straight in to IRIS, as too I presume will Kashflow, from which stat accounts and ixbrl/ct600's are a doodle.
Pricing
@ OP - what is the anticipated pricing for this? Bearing in mind Xero charge say £10 a month for a fully blown package, exporting bank transactions into an Excel csv file is free, presumably as your product simply exports and does same basic grouping of transactions into categories you'd be looking at a few quid a month? If so then perhaps you may get a few customers, for example I have a few artist clients who will never ever use an accounts package but they would like to use something very very simple that doesn't involve completing an Excel template. Would it save me a lot of time? Not a lot, I'd still need to correct a lot of the entries as I believe it is virtually impossible to categorise some expenditure. E.g. That bank payment at Nando's, was that subsistence, entertaining or drawings?
Sounds good
Any tool to ease the pain of bank account analysis would be most welcome. We currently use Autorec for scanning bank statements however a tool like this would make the process even more straight forward. I look forward to hearing more...
Thanks John for the endorsement.
AutoRec is another accounts preparation software solution on the market for anyone looking for a solution.
It is typically used by accountants in practice for VAT analysis for incomplete records or 'shoebox' clients. It’s not for every client but it does save accountants a massive amount of time. AutoRec has been on the market for just 2 years with almost 1,000 accounting practices subscribing to the software.
I know from past research that clients don’t like to give there bank details to any third party? Also AutoRec can also sort by date, transaction details, account code etc. There is a specific export for accounts software systems such as Xero, Sage, VT, IRIS, Digita and many more...
Trial data
What sort of trial data have you been using? If you have only used the program to code your own statements, whose content you can predict, then that isn't a very good test. If you have been testing it on genuine real-world data from a variety of sources then achieving a high level of accuracy becomes more impressive. some transactions will need to be reviewed but so far we have found the particular coding system we have been trialling is very accurate.
I can't help thinking that a full review is still likely to be required. If several suppliers use the same factor, how will those transactions be separated out? Business and personal expenditure may not always be obviously split. Just how big a body of payee names will it recognise? The consequences of HMRC picking up something being miscoded as deductible that isn't are too high to leave it entirely to an automated system.
Demo available?
When you say real world data, do you mean from a variety of sources like I asked? Your own bank statements would be real world data, but would not be a good test as I already pointed out. I also raised a number of problem areas I could foresee, but you haven't really addressed them. A review of client book-keeping is a different matter entirely. Whilst they sometimes make bizarre coding choices, they are at least made by a human who should know exactly what they have bought. That is a very different thing to a program just picking up payee details, which are often not even complete names, from the bank.
Is there a demo available? If not, you really need to put one together. Your claims of accuracy are key to whether this is a worthwhile package. I'd want to be able to test on some data of my choice before accepting those claims at face value. I've already pointed out some issues I expect and others have expressed scepticism in other areas. I'd love to be proved wrong, but I think you will actually need to provide some sort of proof to get sales.
strategy
It sounds as though you're suggesting something to compete with features like 'bank rules' in Xero/Freeagent etc. to some extent. These can be time consuming and produce mixed results, so it's possible this might be a niche that would benefit from attention.
Of the systems I've come across, Money Dashboard stood out for its transaction recognition feature, although designed for the personal banking market and with the added drawback of not allowing user-input in defining the account classifications.
If you could create something to handle this problem well from a business perspective with satisfactory large-scale testing, other software providers might also be interested in it as a module to enhance their existing offerings.