SYSTEM1: Better-Than-Expected FY 2021 Results Reveal Accelerating Data Transition And Bold £1 BILLION Market Cap Opportunity

14 thoughts on “SYSTEM1: Better-Than-Expected FY 2021 Results Reveal Accelerating Data Transition And Bold £1 BILLION Market Cap Opportunity”

1. Hi Maynard,

   Excellent work as always.

   Regards valuation I think you can argue that whilst the the SP was materially lower last year validation of the data products has only more recently occurred – I’d imagine no one who followed the non take up of ad-ratings could have predicted the success of Test my Ad. I’ve been astounded by the data products growth rate – 1%, 7%, 15%, 28%, 38% of rev is staggering QonQ growth imo.

   • Thanks Patrick.

     Yes, I would agree the growth of Data revenue has been astounding following the AdRatings disappointment.

     I think there were signs the business had turned a corner (at least profit wise) and things were moving in the right direction (with ITV/Test Your Ad/UK revenue) within the November interims (and perhaps even in the October update just before). Also, the directorspeak back then was rather bullish for a business that had suffered several mishaps in the prior years. While I did not foresee the rate of Data growth, the valuation last year did not really reflect any potential growth anyway, and the risk/reward situation appeared favourable at the time if you believed trading would one day pick up.

     Maynard

2. Well played on upping your stake so early in the Test My Ad progression.

   I’d held relatively briefly on Ad Ratings potential Nov 18 to Feb 20 selling when it appeared to be gaining no traction.

   Started taking an interest again after the 21st April TS highlighted Test My Ad growth and took a stake when there was a seller in the market in early July (at 225 & 226 with a tiny sliver at 214).

   I’d be interested in how you see the more traditional consultancy business progressing. It appears that whilst the data product is incremental, the consultancy business is declining? Or do you see a churn of consultancy customers moving from consultancy to self serve platform customers?
   In an ideal world I’d like to see the successful data products business driving customers to Sys1 consultancy services but perhaps that’s too much to ask for and Sys1 will evolve to an almost exclusively platform business in time.
   Be interested in your views?

   Patrick

   • Hi Patrick

     I am going on the results narrative, which said:

     “The products shown in the third layer are higher value-add consultancy ‘Improve’ products, which are more standardised than our previous consultancy services. We continue to undertake large, bespoke consulting assignments for a small number of major customers but anticipate that this type of offering will decline in significance for System1 as customers convert to the faster, cheaper standard products

     So revenue per Consultancy assignment will decline as customers convert to the cheaper option. But how much extra Consultancy work will be generated via the lower pricing remains to be seen. The lower charge per assignment could be offset by a much higher volume.

     The plan is indeed to drive greater volume of testing via automated products, which then ought to drive greater Consultancy activity for the follow-up advice. So SYS1 could well evolve into an almost exclusively platform business.

     The question I have is just how ‘scalable’ are the Consultancy products, as some human skill/intervention is needed to give some advice and the associated revenue will to a certain extent be determined by employee productivity.

     I trust your past success with JDG and GAW will now rub off on SYS1 :-)

     Maynard

3. Whilst it’s impossible to accurately predict the future, of my 20 odd stock portfolio SYS1 does seem to have the most obvious attributes of a potential multibagger.
   Scalable platform model, respected operator within their industry and plenty of cash to internally fund development. If it emulates JDG and GAW then we’ll both be in clover :-)

   Another stock I feel is really interesting currently is GDWN. It’s one you’ve previously covered in 2015 & 2017. I visited the company in early September and feel their multi year transformation from reliance on O&G is about to bear fruit.

   Their legacy businesses in Casting and Engineering have secured very long term (multi decade) contracts in nuclear decommissioning and US naval forgings which will ramp up over the next couple of years which have necessitated an expansion of facilities and at Goodwin International O&G valve sales declining from 43% of revenue to less than 33% this year (with valve sales budgeted to be static).

   The refractory business is emerging from covid softness in it’s end markets and has a potential new driver in a silica free casting powders which will allow it to target US casting customers. This business also has 2 new patent protected products that are growing from a low base – SoluForm https://www.soluform.co.uk/about-soluform/ and Lithium Battery Fire Extinguishers https://www.avdfire.com/.

   Management have high hopes for Easat Radar Systems which has evolved from providing components for Radar systems to providing complete systems which they believe will transform it’s prospects.

   Long term perhaps the most exciting business is an internal startup Duvelco which is a company formed to manufacture a polyimide polymer targeting high value components in demanding end markets – aviation etc.

   The most recent results highlight an virtual doubling of EPS in H2 over H1 which I believe could well be set to continue into the forthcoming year as the new contracts continue to ramp up. No forecasts out there to provide guidance but I think they’re probably on a forward PE of <15 with excellent prospects of growth over the next couple of years.

   Perhaps this post shouldn't be residing on the SYS1 report as I seem to have gone off topic …….sorry :-)

   • Thanks for the info Patrick. Will probably have to catch up with Goodwin at some point. Lots has happened since I last looked. Could well be a company for a SharePad write-up.

     (EDIT 20 Oct 2021: And here is the SharePad write-up)

     Maynard

4. Back to SYS1:

   I recently downloaded their Streaming and Entertainment Test Your Brand report for research purposes.

   Anyway today received an email from a Client Development Manager suggesting how SYS1 could help me with my brand recognition :-)

   Suggests they’re being pretty proactive chasing sales leads.

5. Hi Maynard

   Interesting company and their claim to be a future $1b company does make you sit up and take notice.

   Where do come out on the strength of their moat? I have difficulty figuring it out. Is it their predictive power, the speed of analysis or cost advantage? Kantar seems to offer similar products, citing 200k case studies and a similar turnaround time, and they also claim to analyse emotional response, and Kantar is of course bigger. S1 also draws a comparison with Yougov’s business model, but their products seems to be very different. Unlike yougov, What S1 calls ‘Data’ is not really data at all but ad testing. Its data subscription product – Adratings – doesn’t seem to have worked for S1 and I can sort of see why as for an ad, it’s probably more useful to receive tailored feedback than to pick through competitors’ ads, so the recurring high margin subscription revenue may prove elusive.

   Regards,

   Jerry

   • Hi Jerry,

     Ultimately the competitive advantage will boil down to the predictive power of the testing service. The hard part in all of this is correlating the prediction with subsequent changes to brand recognition, market share etc.

     Kantar does offer a very similar service. Kantar interestingly states: “Understand instinctive responses with facial coding (partnering with Affectiva)”

     Which leads on to this:

     “Affectiva, the global leader in Emotion AI, today announced findings on consumers’ emotional engagement with ad content over the last eight years, which show that advertisers are increasingly pulling on people’s heartstrings to generate a response.”

     “This first-of-its-kind emotion database consists of data on 10 million consumer responses to more than 53,000 ads in 90 countries, collected over the last eight years. The data was collected through Affectiva’s Media Analytics software, which uses artificial intelligence to detect nuanced emotional reactions and complex cognitive states from the facial expressions of market research participants when they view ad content.”

     Affectiva uses facial recognition, unlike SYS1, which uses simple user responses. I don’t know which is the more accurate option. Other competition will exist, too.

     Although SYS1 claims its services are better, cheaper and faster, I recognise the company may not (and never will) have a Google-type dominance. That perhaps has been reflected in the past valuation. If/when the rating starts to rise to premium levels, the size of the moat will become more important to investors.

     SYS1’s founder remarked at the AGM the other year of wanting AdRatings to become the “Yougov of advertising data“. The AdRatings database has not worked out as a standalone product and I doubt subscription revenue will ever be material. But the data helps put new predictions into context for the client, and can be used for the all-important correlation between past adverts and subsequent market-share gains.

     Maynard
     PS I am hopeful the ITV partnership validates SYS1’s standing among effective ad testers.

6. System1 (SYS1)

   Trading update published 20 October 2021

   An encouraging update. Here is the full text interspersed with my comments:

   ——————————————————————————————————————
   System1 Group issues the following update on trading for the six months to 30 September 2021. The Company will announce its interim results on 30 November 2021.

   The first half of the current year (H1) saw a marked increase in sales of automated data products including Test Your Ad. Data products represented 45% of Revenue in Q2, growing from 28% in Q1. Data Revenue in Q2 exceeded Q1 despite a quarter-on-quarter reduction in total Group Revenue. The Group’s H1 Revenue grew 22% on last year to £12.3m.
   ——————————————————————————————————————

   Q2 Data revenue was better than the Q1/AGM statement had indicated.

   Total H1 revenue of £12.3m less £6.5m for Q1 gives £5.8m for total Q2 revenue. 45% of that £5.8m gives Q2 Data revenue of £2.6m and H1 Data revenue of £4.4m — representing 36% of total H1 revenue. SYS1 had previously indicated Data would represent a third of H1 revenue.

   Total Q2 revenue of £5.8m is not great given the previous two quarters each registered £6.5m total revenue. But SYS1 is transitioning from bespoke consultancy work to automated Data income and I guess the revenue cross-over will not always balance out. What is important is the more attractive Data revenue is growing quickly:

   

   Consultancy revenue was only £3.2m during Q2 versus £4.7m during Q1 and £5m-plus for Q3 and Q4 of FY 2021. That reduction has meant current profit is running below what I had calculated in the blog post above.

   ——————————————————————————————————————
   Adjusted Operating Costs* grew as planned as the Group continued to invest in automated products, technology and business development, increasing by 10% over the comparable period to £9.0m.

   Adjusted Pre-tax Profits are expected to be some £1.3m in H1, approximately £0.9m higher than in the comparable period. Statutory Pre-tax Profits are expected to be £1.3m compared with a loss of £0.4m in the first half of last financial year.

   Period end cash, net of borrowings, was £7.3m, compared with £6.5m at end-March 2021.
   ——————————————————————————————————————

   Adjusted operating costs of £9.0m for this H1 is a little better than I had predicted. My sums in the blog post above predicted nearly £9.2m.

   Pre-tax profit of £1.3m for this H1 looks about right assuming an 85% gross margin (as per Q4 FY 2021) on the £12.3m H1 revenue (=£10.5m gross profit) less £9.0m of adjusted costs.

   But H1 pre-tax profit of £1.3m, or £2.6m annualised, is below my £3.6m full-year guess within the blog post above:

   

   I am hopeful this Q2 is SYS1’s profit low point as Data revenue continues to grow and replace the old-style Consultancy work. H2 could provide some clues as to the possible economics of the Data operation, as any incremental revenue/profit during Q3 and Q4 can be compared to that of Q2.

   Net cash of £7.3m is up only £0.8m on the year end, which seems a tad light given H1 pre-tax profit of £1.3m. But the Q1 update implied net cash was £7.6m, which we now know to be £0.3m higher than Q2, so SYS1’s cash movements appear significant from quarter to quarter.

   ——————————————————————————————————————
   We have been delighted by the continuing adoption by both new and existing customers of System1’s repeatable, fast-turnaround and scalable data products as they displace the historic large bespoke consultancy projects that dominated the Group’s activity until H2 last year. We believe that the growth in data revenues signals tangible progress towards our strategic goals, and that these revenues will in due course more than offset the reduction in bespoke consultancy projects as more customers switch to using our automated data products.

   In line with previous communications, the Group intends after the interim results announcement to initiate a share buyback programme to repurchase its shares over an extended period, in order to enhance shareholder returns and to satisfy obligations in relation to employee share schemes.

   * Adjusted Operating Costs exclude impairment, interest, share based payments, bonuses, severance costs and government support related to the Covid pandemic. Adjusted figures exclude items, positive and negative, that impede easy understanding of underlying performance.
   ——————————————————————————————————————

   Ah, so SYS1 is indeed expecting Data revenue to “more than offset” Consultancy work, albeit “in due course“.

   The mooted buyback is still on, although SYS1 is now mentioning an “extended period” and “obligations” to share schemes. So not quite the ‘pure’ buyback I had hoped for.

   (There are of course no buyback “obligations” with share schemes; companies do not have to operate such schemes and do not have to buy back the shares to offset the resultant dilution.)

   I am now reluctant to consider the buyback for valuation purposes as the text reads as if the money will be spent over many years to keep a lid on share dilution. Mind you, significant option vesting will only occur if revenue doubles to £45m during the next four years — which if it does ought to support a higher share price anyway.

   Maynard

7. Agree with that Maynard, encouraging progress with Data Products (still presumably only Test my Ad with the remaining 2 pillars still to be reflected in the numbers) offset by alarger than hoped reduction in Q2 consultancy revenue.

   He’s an example of their work with GSK which I thought quite interesting as it highlights the iterative creative process that SYS1 can assist with.

   https://www.adweek.com/performance-marketing/how-glaxosmithklein-used-emotion-tracking-to-hone-its-marketing-for-arthritis-drug/

8. System1 (SYS1)

   US trade name legal dispute

   I have been tipped off about a legal dispute:

   A US company called System1 Inc is about to float on the NYSE. Here is the S-4 flotation prospectus.

   The relevant legal text:

   “We have received correspondence from counsel for a United Kingdom-based marketing research company and its United States subsidiary (collectively, the “Demanding Group”) alleging trademark infringement based on our use of the “SYSTEM1” trade name and mark in the United States, and alleged use of the “SYSTEM1” trade name and mark in the United Kingdom.

   The correspondence demanded that we cease and desist from using the “SYSTEM1” name and mark, and made reference to potential legal action if we do not comply with that demand.

   While we were engaged in active discussions and correspondence with the Demanding Group to resolve the matter, the Demanding Group filed a lawsuit in the United States District Court for the Southern District of New York on September 27, 2021 (the “Infringement Suit”) alleging (i) trademark infringement, (ii) false designation of origin, (iii) unfair competition and (iv) certain violations of New York business laws, seeking, among other things, an injunction, disgorgement of profits, actual damages and attorneys’ fees and costs.

   We believe that the Demanding Group’s infringement and other allegations and claims set forth in the Infringement Suit may be subject to a laches defense, among other defenses, and we intend to vigorously defend our rights in the Infringement Suit.

   No lawsuit has been filed in the United Kingdom, and we do not believe that our activities infringe any rights of the Demanding Group in the United Kingdom because, among other defenses, we do not offer services to customers using the SYSTEM1 name and mark in the United Kingdom.”

   I am not a legal expert, but SYS1 may well have a good case having operated under the System1 brand (in the US and globally) since 2017.

   Although System1 Inc does own the https://system1.com/ domain, and I am pretty sure SYS1 would have noted the ownership of that domain before plumping for https://system1group.com/.

   System1 Inc will not be short of funds following is US flotation (presentation here). The company has already paid $55m to cover transaction fees and expenses:

   

   So perhaps System1 Inc will pay a handsome amount to SYS1 to settle the case?

   Although that outcome may not be ideal if two ‘System1’ companies exist in the States involved in digital marketing in some way. System1’s business model is very different to that of SYS1 though:

   

   The flipside is System1 Inc has a lot more money than SYS1 to contest a protracted legal battle.

   Another issue is whether ‘System1’ is actually SYS1’s to own, as the term was devised/popularised by Daniel Kahneman.

   Maynard

   • System1 (SYS1)

     Complaint for trademark infringement published 09 November 2021

     SYS1 claims the case mentioned in the comment above is “compelling“. Here is the full text:

     ——————————————————————————————————-
     “System1 announces that on 27 September 2021 its US and UK subsidiaries, System1 Research Inc. and System1 Research Limited, filed a complaint for trademark infringement, unfair competition and deceptive trade practices at the United States District Court Southern District of New York against System1 LLC (“the Opposing Party”) over their infringing use of the mark “SYSTEM1”. The Company is being advised in this matter by Norton Rose Fulbright, an international law firm.

     The Opposing Party continues to use the “SYSTEM1” name and mark and has announced that it proposes to expand such use by becoming a NYSE-listed company through a business combination with Trebia Acquisition Corp.

     The Company takes the protection of its unique intellectual property very seriously and believes such ongoing and expanding use by the Opposing Party is likely to cause significant confusion with System1 Group PLC and its registered marks and, moreover, constitutes intentional infringement, in violation of relevant US federal and state trademark laws as well as trademark laws in the UK, the EU and other countries.

     The Board believes that the Company’s case against the Opposing Party is compelling and will issue further updates in due course. A summary of the Company’s registered trademarks is attached to this release.”
     ——————————————————————————————————-

     Here are the trademarks registered in the US:

     U.S. Registration No. 5,362,152 for “SYSTEM1 RESEARCH”
     U.S. Registration No. 5,488,953 for “SYSTEM1 GROUP”
     U.S. Registration No. 5,870,379 for “SYSTEM1 MARKETING”
     U.S. Registration No. 5,870,950 for “SYSTEM1 AD RATINGS”

     So nothing for “SYSTEM1” in the US, even though SYS1 does have such trademarks in the UK and EU:

     EU Registration No. 018400922 for “SYSTEM1”
     UK Registration No. 00003596520 for “SYSTEM1”

     The obvious question then is who owns the “SYSTEM1” trademark in the US?

     I can’t quite see the “compelling” case myself, and perhaps the best outcome for SYS1 is System1 Inc pays off SYS1 handsomely to go away.

     Maynard

9. System1 (SYS1)

   Publication of 2021 annual report

   1) Section 172

   New for the 2021 report is a full Section 172. The s172 for the previous year just referred readers to other parts of the report. Within this year’s s172…

   A nice pyramid reiterating the company’s bold market-leading ambition:

   

   Reiterating “better, faster and cheaper“:

   

   Useful snippets about LinkedIn and how to devise an advert, with a start, middle and end:

   

   ‘Truth’ and ‘trust’ is important to the company:

   

   ‘Truth’ always helpful when adverse trading updates are published! Interesting to see the company applies its own service for employee surveys. Presumably the results are accurate.

   Snippets about shareholder alignment:

   

   Contrary perhaps to the text, the online 2021 AGM did not allow investor participation on the day (questions had to be submitted in advance). Whether the AGM had to be online was questionable, too. The event was hosted in August, the same month I attended a physical AGM for a different company.

   Useful to know there are no plans to raise extra cash!

   Interesting insights about the suppliers:

   

   All seem to be on-board for helping SYS1 disrupt the industry.

   2) Risks

   No major changes and only the one extra ‘sub-risk’.

   A few bits of revised text for this year, including:

   

   For 2020 the mitigation text for losing a significant customer included: “We therefore go to considerable lengths to monitor service quality and seek client feedback.” Now SYS1 just “works hard to earn their loyalty“.

   Revenue from the largest customer at 8.1% equates to £1,861k, versus £2,596k for 2020. Never clear whether SYS1’s largest customer is always the same client. But the contribution has bobbed up and down over the years, ranging from £1,476k for 2012 to £3,402k for 2017.

   The extra sub-risk relates to the company’s transition to automated products:

   

   3) Corporate governance

   Looks like the directors all attended their relevant meetings:

   

   Mr Barden and Mr Willford attended two main board meetings before becoming formal executives in June. Also, the three executives sat in on the rem-com meetings, which is notable because Mr Kearon enjoyed a useful 11% pay hike despite the company collecting furlough payments and him delegating some duties to Mr Barden (see blog post above).

   I recall the 2020 report had confirmed no execs were involved with their own remuneration:

   

   But this 2021 report removed all that text:

   

   Not sure why somebody removed that text. Surely Mr Kearon did not have a say in deciding his extra 11% pay hike?!

   A review of the board and chairman (new for 2021) did not reveal anything untoward:

   

   4) Audit and audit report

   The non-exec in charge of the audit committee is in fact in charge of three other audit committees, so not unfamiliar with auditors:

   

   The new FD decided to change the accounting software:

   

   No audit problems:

   

   Slight change to key audit matters. Just the sabbatical provision looked at this year (and not intangible asset valuation as well). The impairment of the intercompany receivable was added for 2021, but is somewhat academic from a group perspective:

   

   Going-concern audit text all revamped for 2021, including notes on worst-case scenarios and covenant breaches:

   

   Reads as is SYS1 would still be able to pay back the bank loan if big trouble strikes.

   Standard 5% of pre-tax profit for audit materiality:

   

   100% audit coverage, but a full audit was performed only in the UK — which represents just 30% of revenue:

   

   Mind you, the previous auditor performed only audit ‘procedures’ on the subsidiaries for 2019, so a full audit at the UK subsidiary is an improvement on that.

   New standard audit text for 2021 covering fraud:

   

   5) Director remuneration

   A reminder that the LTIP participants do not receive an annual bonus:

   

   The remuneration policy says revisions should happen every four years…

   

   …but then lists a load of revisions for the LTIP (see blog post above):

   

   “Implementation of the new strategy is unlikely to be linear…” has (sadly!) been confirmed by subsequent trading statements:

   

   The LTIP was changed because management could have been “disincentivised“:

   

   Reads as if management would have actually been discouraged to perform without the revised LTIP, which does not sound great.

   06) Going concern

   Revamped text for 2021:

   

   “the Group is a lot more confident about how to respond to an abrupt negative situation” should be useful given subsequent trading updates!

   07) Revenue recognition

   Nothing untoward. Revenue booked when the consultancy report is delivered:

   

   08) Customer and supplier payments

   Useful snippets on payment terms. Customers should pay within 60 days of delivery, and some customers pay upfront:

   

   Trade receivables were £5.3m, which versus revenue of £22.8m suggests payments take (5.3/22.8*365) = 85 days:

   

   Doubling up the stronger H2 gives revenue of £25.4m and implies payment terms of 76 days. So customers are can be tardy payers.

   Footnotes indicate 33% of trade receivables were overdue (£1.7m/£5.3m), which is not ideal but not out of keeping with previous years:

   

   Overdue trade receivables were 29%, 36% and 34% of total trade receivables for 2020, 2019 and 2018 respectively.

   Working-capital cash movements have never really appeared untoward, so I guess SYS1 eventually collects these payments.

   The largest customer is a slow payer: 



   

   Largest customer represented revenue of £1,861k, and receivables of £666k indicate payment terms of 666/1861*365 = 131 days.

   Trade payables remain relatively small at £845k, with terms quoted at 30-45 days:

   

   The level of trade payables is not great when trade receivables are 4-5x larger (i.e. SYS1 pays suppliers much earlier than customers pay SYS1).

   But SYS1 does carry notable accruals and deferred income, which I suspect are mostly expenses recognised but not yet invoiced by the supplier. So the overall receivable/payable cash flow movement may not be so pronounced.

   Also, SYS1 does benefit from contract liabilities, which reflect upfront customer payments and are double the size of ‘contract assets’ (external costs capitalised for project work not yet completed):

   

   09) Borrowings

   Interest charged on the £2.5m loan is a reasonable LIBOR+2.5%:

   

   Finance lease IFRS 16 obligations reduced to £2.6m from £4.3m for the previous year after office closures:

   

   The associated right-of-use asset of £1.3m implies SYS1’s obligation of £2.6m is twice as much as would be the case if the remaining lease agreements were started today.

   10) Options

   Confirmation of the option count being 12.8% of the year-end share count (adjusted for treasury shares). Most options concern the LTIP:

   

   11) Employees

   New for 2021 are helpful entries concerning R&D:

   

   Total R&D up 20%.

   Staff numbers down in all areas save for IT, which understandably increased given the shift towards Data revenue:

   

   The lower headcount led to lower employee costs:

   

   Nothing truly shining through on the workforce productivity front to emphasise the Data transition. Revenue per employee was £170k, a fraction below the average of the preceding five years of £177k. Total cost per employee was £79k, just about matching the preceding five-year average of £80k.


   Not something awarded at every company: staff earn an extra 20 days holiday after serving for six years:

   

   12) Tax

   Confirmation of R&D tax credits, which if continued may prove helpful for cash flow purposes:

   

   Maynard

Comments are closed.