A lot of the talk is on SRED credits. SRED is a large part of the credit program for startups in Canada. Though it's not as simple as "Canada pays 80% of your salaries"
You can get up to (off the top of my head) 75% of salaries paid back on hours worked "advancing technology". That means it's only for true R&D. That means UI, business logic, API, App anything, expenses, costs, etc... don't count. It's only for when you embark on a project that truly advances the worlds state of technology where you can really reap the benefits of SRED.
Most startups, as sad as it may be, don't truly advance technology. A lot create value, yes, but advance technology? not really. Certainly not all year. So they'll only get 75% of salaries paid for the hours they worked on advancing technology.
You'll get that back at the end of the year as one lump sum. It's awesome. However, it goes to the company and the company has to have already funded the salaries -- it wont fund salaries, it will reimburse them.
On top of that the system is so hard to navigate that you'll end up requiring a consultant to navigate it. They'll take a cut, between 10-20%, of the return.
It's not as simple as the article states, and our poutine doesn't look like that.
As someone who has actually used the SR&ED program for many years, I have to say that you've got it a bit wrong...
SR&ED will cover a percentage of capital expenditures essential to the work and a percentage of $ paid to contractors so it's salary plus SOME eligible expenses. Salaries make up the biggest part though. I got about 65% back, using the proxy method - a simpler way of calculating SR&ED overhead on eligible salary expenses.
The R&D effort does not have to advance "the world's" state of technology, just your company's.
It can't be obvious or routine engineering BUT work that is required to complete the R&D work (but not directly R&D) IS eligible and you can claim it.
The 3 core requirements of SR&ED eligibility (for software anyway) are (1) there has to be technical uncertainty, (2) you must use a systematic approach and keep records a notes and (3) you must achieve a technological advancement.
The 3rd item catches people because it doesn't require a successful outcome. You can fail and still claim the project. If you knew in advance (or it was obvious to anyone in the field) that it would fail (or succeed) then there was no uncertainty and it would not qualify.
SR&ED is really not that difficult to navigate. Consultants can help, but they'll take a big chunk of your claim. Most of them work for a percentage that runs anywhere from 15% to 35%. That's big money for very little value. I wrote my own claims for many years and they were all successful.
When I started there was no limit on the write ups, now they put some severe character count limits on the writes ups. That makes them harder for us but less confusing for the reviewers. You really have to be able to distill things down to just the essentials.
Any Canadian software company that doesn't at least look at SR&ED is just throwing money out the window.
They've recently removed the ability to claim capital equipment. My experience, which is probably less than yours, is that the R&D requirement is real and omnipresent - anything that's spent on production or that will go into production can't be claimed. Furthermore I've been asked for lab notebooks and experiment designs to justify the materials I used that I claimed were for R&D purposes.
They used to have an "Account Executive" program where you could talk to a reviewer BEFORE you started work to explain what you were doing and get tips on what was eligible and what was not. I did that when I started and it was an eye opener!
The reviewer I spoke to was the one who handled all our claims, so it didn't hurt that he already knew about us. It made his job, and mine, a lot easier.
Wait... okay, found a link to the services they offer and Account Executive is still in there, plus a bunch of others...
If you're at all uncertain about what you can/should be claiming then you should definitely talk to a SR&ED professional or use one of these services to get definitive answers from the source.
In theory SRED is supposed to be for R&D. You'd be surprised what can count for R&D. Doing A/B testing to solve the problem of which UI is more preferred/efficient? That can count for SRED work.
What works in practice wouldn't be held up in an audit. Depending on the mood of your approving agent they can deny the whole application if you mention specific interface words.
You'd be better off creating and Irap project to cover UI and business logic items. IRAP is out of the same government fund but pays slightly less for a project as you go, not one lump at the end of the year.
IRAP and SRED are mutually exclusive but if you spend time thinking about it you'd be able to decide work to fit into both nicely.
I had to go through the whole SRED song and dance at my previous $work. The company writing up the proposal was really eager to get the numbers high, but the when the government agent came down to the office to go over the expenses the whole experience was so painful that the owners of company decide to pass on even applying for SRED in subsequent years.
Similarly at a small startup I worked at were surprised at how rough the auditors were with us. We certainly thought we were 100% in the clear and doing genuinely innovative work but they really pressed us on it. Getting SRED is not easy as some say.
This shouldn't dissuade anybody from trying. That's the whole point of business. Going out on a limb. Read the rules, follow them, and if it doesn't work out... at least you tried.
I agree. A key part not mentioned is that you have to fund the salaries before and it's only paid much later! As in after you close your fiscal you then have to submit and it takes about a year before you're paid. So in total you have to cover the salaries for two 2+ years before you're paid. You also have no guarantee you will be paid.
It's also only on developer costs that are accepted and only on R&D work: none of the business costs etc are covered. In many cases these are just as significant or more.
There's also a reason that a consulting industry has started around it, it's a good amount of effort. Again helping more advanced companies as they can afford the costs for the 2+ year it takes before you can get it back. And again nothing is guaranteed.
Essentially it's not good for startups or smaller companies, it really only benefits companies of a medium size that were alreday going to cover this expenses by reducing some of their costs to increase them growth rate. But for a small company, less than 10 employees it's really hard to actually take advantage of the program in real life.
Plug: the UK govt also has R&D tax credits, and it's what my company, GrantTree, specialises in.
They're similar to Canadian tax credits, but can give you up to 25-32% cash back, and are a fair bit more lenient on what passes audits (at least compared to the horror stories I've heard from Canada).
Feel free to get in touch with me to find out if/how you can get some of that in the UK.
Are UK salaries even worse than Canadian ones? I've looked at analytics related positions in the UK and was shocked at how bad the average pay seemed to be given what the cost of living was.
It seemed to be the same for software developers.
Given what patio11 says about salaries in Japan, that country also sounds absolutely terrible.
It's actually 80% of the salary for any work that has been done in R&D, but like you said, most of the work isn't.
A few years ago a lot of stuff could be considered as work "between two uncertainties", including non-engineering work such as graphic design, but the rules are much more strict nowadays. I think they still give credits for things that are way too simple to be R&D (ie: normal problem solving), but we've been doing [small] claims for the past 3 years and the rules have been tightened a lot.
Also, doing a claim requires you to provide evidences that you were following a certain R&D methodology. It's definitely worth it, but it's a pain.
Yeah, they immediately start parsing and backpedaling the headline: "the Canadian government is paying almost 80% of his <developers>' salaries.". Still disapointing they don't delve into the actual program structure though.
You can get up to (off the top of my head) 75% of salaries paid back on hours worked "advancing technology". That means it's only for true R&D. That means UI, business logic, API, App anything, expenses, costs, etc... don't count. It's only for when you embark on a project that truly advances the worlds state of technology where you can really reap the benefits of SRED.
Most startups, as sad as it may be, don't truly advance technology. A lot create value, yes, but advance technology? not really. Certainly not all year. So they'll only get 75% of salaries paid for the hours they worked on advancing technology.
You'll get that back at the end of the year as one lump sum. It's awesome. However, it goes to the company and the company has to have already funded the salaries -- it wont fund salaries, it will reimburse them.
On top of that the system is so hard to navigate that you'll end up requiring a consultant to navigate it. They'll take a cut, between 10-20%, of the return.
It's not as simple as the article states, and our poutine doesn't look like that.