appScatter — An innovative play on the app market

appScatter is an early-stage B2B mobile app distribution, management and intelligence company. Its cloud-based platform helps its customers (app developers and publishers) to publish to and monitor more than 50 app stores globally, with more being added. Beyond the key Google and Apple app stores, the global app store market is fragmented and monitoring performance on an individual app store basis can be burdensome. appScatter’s SaaS-based platform enables customers to register each app once, for multiple distribution, thereby overcoming the technical challenges of distributing, updating and monitoring their apps across multiple stores. With its data mining capabilities, this is enabling the company to build one of the world’s largest repositories of multi-store app ranking data, which it plans to monetise from next year. The planned launch of a marketplace for add-ons in Q118 is another potential source of revenue. The group listed on AIM in September 2017, raising a net £7.4m, and the public launch of the platform was on 22 November.

appScatter began to license its platform on a restricted basis in January 2017. Maiden revenues of £0.9m were reported in H117, and as of June annualised recurring revenues (ARR) were £1.7m, providing a starting point for forecasts. The group has also built a strong pipeline of potential licensees; 10,000 accounts have registered an interest in using the platform (on a free basis) and management anticipates a high conversion rate to paying licensees post the public launch of the platform. In addition it has developed an extensive database of targets. Following launch, we expect the company to ramp up its marketing efforts, and widen its network of strategic partners.

Given the limited history, we base our forecasts on the group’s experience to date but factor in an element of caution regarding marketing effectiveness, price discounting and the cost base. We assume from 2018 that the cost per acquisition (CPA) is 50% above that experienced in test marketing and that the conversion rate (ie from accounts that have registered an interest in the platform to licensees) drops from 4% to 2% after four months. In 2019, we assume that marketing spend is increased even further and to support forecast rapid growth we assume an increase in headcount from 33 to 61 and a 20% increase in other costs. The business model should be highly scalable and, due to this high degree of operational leverage, we expect the company to achieve operating margin break-even during H119, with strong operating profitability by H219 (22% margin). Although this rapid progression towards profitability is unusual among SaaS-based peers, we note that appScatter’s profile is most closely aligned to Atlassian (see page 19), which shares a similar cost structure. We summarise the key drivers underpinning our forecasts in Exhibit 1 below. Given the early stage of the company, we also demonstrate the impact on our forecast should we flex certain key assumptions: price (for instance if volume discounts are offered) and conversion rates (3% rather than 2%).

Based on these forecasts, appScatter trades on an FY18 EV/sales multiple of 4.1x, similar to its larger (more mature) SaaS peer set, although on a premium to other UK software groups. With the strong forecast revenue growth, this premium would convert to a significant discount to SaaS peers in FY19 when it trades on 2.0x sales (peers 5.1x) and 15.1x EV/EBITDA (peers 35.1x).

The valuation opportunity hinges on the successful commercial launch of its platform. As a first mover in this area, there is currently little visibility regarding the level of demand for its services. However, 10,000 accounts have registered an interest and visibility on some of the key variables will start to improve as the group moves into 2018.

We have tried to capture early-stage uncertainties by using a higher CPA than that experienced to date by management, factoring in an element of price discounting and assuming a 20% increase in the cost base in 2019. We also use a relatively high 15% discount rate in our DCF and assume that after the initial forecast period, revenue growth recedes to a CAGR of c 20% to 2027 (and 2% in perpetuity). This returns a value per share of 67p. On evidence of the demonstrable success of the platform, we would expect the discount rate to come down; we generally use a WACC of 12.5% for technology groups that have demonstrated demand but are still loss making, which would return a DCF of 89p. Investors should also consider:

Given the early-stage nature of the business there is a higher than average degree of sensitivity to our forecasts, particularly in relation to the pace of adoption, the cost to acquire customers, the rate of conversion to paying customers, renewal rates and the price of the service. Other risks include potential competition in the future, as well as currency risks, as appScatter is planning to switch its pricing from sterling to US dollars. Should the business not proceed as anticipated, appScatter may require further funding to achieve cash flow break-even.

appScatter is a B2B mobile app distribution, management and intelligence company. Its SaaS-based platform helps its customers, which include app developers and app publishers, overcome the technical and fragmentation challenges of easily distributing, updating and monitoring their apps across more than 50 app stores globally.

The platform is distinct from other offerings that tend to focus on the two largest app stores, Apple’s App Store (iOS) and Google Play (Android). From a single point of registration, appScatter’s platform offers the widest reach in terms of automated app store registration and distribution. With its data mining capabilities, this is enabling the company to build one of the world’s largest repositories of multi-store app ranking data. Across the platform it tracks c 977m unique app URLs from 8.6m apps and monitors approximately 2.1m app publishers. This data could provide valuable intelligence to companies regarding their, and their competitors’, app performance and in addition to licensing the platform, management plans to start to sell this data.

Management restricted access to the paid services on the platform in anticipation of the public launch, which was on 22 November. As at 27 September 2017, c 10,000 publishers had registered an interest in using the platform. At 31 July 2017 it reported over 800 paying users with over 3,000 apps on the platform. Customers range from individual developers to large enterprises and some significant customers include Allianz, City AM, Emirates and Lab Cave.

In Q118, the group also plans to launch a marketplace for add-on services that will enable its users to integrate a wide range of third-party solutions such as in-app analytics, optimisation and monetisation solutions or revenue factoring services (among others). As well as making the appScatter product stickier, this has the potential to provide an additional source of income for the group.

CEO Philip Marcella founded the business in 2013. Philip, who is an experienced programmer, has worked in app development since 2002, running a team that has built hundreds of apps across a range of markets, including utilities, games, children’s books and augmented reality. Concurrently, he recognised the time-consuming nature of app publishing and his team began to build a tool to automate the process and which would also provide clients with access to consolidated sales data across multiple app stores. This same team has been instrumental in the development of the appScatter platform.

In 2013, appScatter LLC was formed in the US, where it began researching and testing apps on alternative app stores. The company recruited its first engineers in 2014. In 2015, the company created a proof of concept for submitting apps and gathering download data across multiple app stores, thereby harvesting data from millions of apps across multiple app stores. appScatter Limited (UK) was formed in 2015, and it acquired appScatter LLC in May 2016.

In 2016, the appScatter platform underwent beta and usability testing, and a sales and marketing team was created. The company introduced the platform at the Berlin app summit in November 2016.

Prior to IPO, approximately £12.6m had been invested in the business, largely through several private placements. On 5 September, appScatter listed on AIM, raising £9.0m gross proceeds at 65p per share (£7.4m net of IPO and fund-raising costs). Post the IPO, Philip Marcella owns 26% of the shares and the company has a 64% free-float.

The intention is to use the proceeds of the IPO for working capital (£4.8m), in particular to fund the growth of the licensed user base, as well as platform investment (£1.7m around the launch of the marketplace) and sales and marketing (£1.3m).

Until recently, the platform was only available on a restricted basis, with some services offered for free and a wider range of services only available to a select number of licensees. The public launch of the platform has added a number of additional features aimed at improving the user experience and increasing subscriber retention.

The appScatter cloud-based platform, which runs across more than 600 servers, is integrated into and collects publicly available data from 50 of the world’s largest app stores (out of a total of approximately 300). It is now tracking more than 8.6m mobile apps and 977m individual app URLs1 from approximately 2.1m publishers worldwide. The platform integrates a telemetry system to power the backend. Data tracked includes revenues, downloads, ranking, territories covered, meta data, pricing and app-specific details including publisher names.

In terms of platform infrastructure, Amazon Web Services is used for data storage and MongoDB for data sets. Riak Time Series databases are used for app ranking data and Amazon Elastic Search is used in enabling appScatter’s search engine.

The core services currently offered include app store distribution, app store rankings and competitor intelligence.

App store directory and search engine: The appScatter directory provides details on each of the 50 app stores that it supports. This includes descriptions of registration and submission processes, reporting procedures, territories covered, size and in-app billing capabilities. Using the appScatter search engine, users can access details on apps and publishers, including ranking position, territories covered, all metadata, pricing and app specific information such as publisher name.

App store registration and distribution: Once registered with appScatter, subscribers have a single point of access to app stores worldwide. appScatter checks for an app’s compatibility with its supported stores and enables its customers to choose to publish apps either by country, device type, language or store. Where an app is incompatible with an app store, the platform will notify the user with the specific reasons such as lack of a brief description or insufficient graphics (Exhibit 2). appScatter currently distributes to 50 stores worldwide, avoiding the requirement for individual registration on each app store. The company is adding four to five stores each month and expects that it will level out at c 100 stores (ie those that are the most efficient and productive for customers to use).

Analytics and ranking: Users can monitor their own app’s sales or download performance by store (which in turn can drive their marketing strategy). It also enables users to quickly check, in one place, an app’s territorial compliance, brand consistency, etc. Performance reports can be segmented by geography, app store or device type. This information can either be accessed via the appScatter dashboard, or via integrated third-party analytics tools (eg Google analytics).

Intelligence: Users can create a customised watch list of competitor or interesting apps or segments in order to monitor market share or other industry trends.

Monetisation: The recently launched appScatter Android software development kit (SDK) enables app developers to conveniently integrate in-app billing to all Android stores integrated on its platform: for instance, Amazon, Bemobi, Samsung and Google Play. This enables in-app purchasing across all app stores, including those that do not currently support this form of monetisation. Where a store does not specify its own in-app billing method, appScatter’s own in-app billing will be used as default. This should appeal to app developers as it has the potential to extend the monetisation of their apps. Unique to appScatter, this is an open-source initiative that management believes will support the wider use and awareness of the appScatter platform.

The platform is available to use for free on a limited basis, with a wider range of services available to licensed users. appScatter segments its licensees into three categories – small, medium and large – with the categories differentiated as follows:

In keeping with other SaaS companies, appScatter charges its customers a regular subscription fee to use the platform. The charge is based on both the number of users ($12 per ‘seat’ per month) and the number of managed apps ($60 per app per month). The minimum sign-up is for a small client account, which would be charged as five users and one app for a total cost of $120 per month. Additional monetisation could occur through the marketplace of integrated third-party software, though the pricing of these products will vary, and in many cases will not generate incremental revenues but instead should boost retention.

With the proceeds of the fund-raise management plans to increase its marketing spend to drive user acquisition, as well as hiring to support user growth and continue developing the platform.

Test marketing began in November 2016 to determine customer cost per acquisition (CPA) over social media, Google AdWords and select conferences such as QCon and App Promotion Summits. Social media and Google AdWords tend to attract smaller customers, whereas the conferences are more fertile ground for developers and enterprise customers. To accelerate user acquisitions, appScatter is also looking to form strategic partnerships with other analytics and mobile advertising networks that serve a similar customer base. It is in discussions with a number of potential partners, whose own users would be able to access their data through appScatter. In some cases, appScatter would pay a commission on revenues generated by these partners.

As of 31 July, appScatter had c 150 paying licensees (which equate to c 800 users or seats managing c 3,000 apps). In addition, there are currently approximately 10,000 users that have registered an interest in the service. Following the launch of the platform, the focus is on converting the existing registered users to licensees. Management has indicated that there is a degree of ‘pent up demand’ from the existing registered user base and consequently in the first few months post launch we would hope to see a steady increase in new licensees to the platform. In addition appScatter has a database of approximately 1.7m names, which will be used to target potential users of its platform over the medium term.

appScatter currently has three employees in the US and the remainder in the UK. On 1 July, the company opened an office in Berlin, initially starting with two developers from its outsourced provider. In time, the company would like to open an office in China to access the domestic market.

The company has plans to increase and improve the services offered on the platform, including:

Marketplace: In Q118 the company plans to open a marketplace where users can directly integrate with other services such as Adobe Analytics, Airpush, JIRA, Flurry, Salesforce, Slack, Xero, etc. By integrating with a wide range of apps, management believes that it will increase the stickiness of the core service. For instance, users could integrate their own ad-networks enabling them to monitor in-app behaviour, return on marketing investment per app store, in-app purchases, etc, to get a fuller picture of the effectiveness of an app’s marketing campaign on a store-by-store or market basis. Another add-on under consideration relates to the factoring of advertising revenues, which could appeal to publishers generating a significant share of advertising revenues from Asia where payment cycles can be protracted. In total, management plans to introduce approximately 300 third-party add-ons over the next 12 months and will target marketing at existing users based on their individual profiles. Add-ons will include a wide range of free and paid for services and over time could provide a valuable additional revenue stream for the group. For example, enterprise software company Atlassian (workflow management and collaboration tools), generates c 6% of its total revenues (c $37m last year – stated net of 75% commissions to providers) from the 3,000 or so add-ons that are available in its marketplace.

Data API: appScatter is becoming the world’s largest repository of app store ranking data and in early 2018 it plans to launch a data API. This will enable customers or data companies to access the vast amounts of publisher data it collects. This data could prove very valuable for a company’s marketing intelligence. Initially it plans to sell the raw data, which is likely to appeal to its larger enterprise customers or data companies that have their own business intelligence (BI) tools. Further down the line, it plans to develop its own BI interface, which would make this data accessible for all of its customers.

Telemetry services: Telemetry functionality is already integrated into the back end of the platform, to enable the rapid access and sorting of data. Telemetry functionality is being extended early next year to further improve the dashboard analytics capabilities for its customers (eg to enable advertising and sales overlay analysis). Ultimately management is working towards the full integration of a telemetry offering into the apps for both frontend (eg number of downloads, revenues) and backend analytics (eg load times, memory usage).

App reviews and translation: The platform is being translated into several languages, with the initial focus on German, followed by Chinese.

appScatter’s solution is relevant for any company that needs to publish apps. This can range from companies where the app is their primary source of income (for instance games developers, entertainment services), or companies where apps are provided to customers to support their wider business case (eg retail companies, restaurants, financial services groups, etc).

The market for mobile goods and services is already a vast market and is forecast to continue to grow rapidly. App downloads are ultimately driven by increased smartphone usage, itself a function of continued improvements in connectivity as operators move towards universal 4G coverage, the availability of affordable handsets (particularly in the fast-growing developing markets), and improvements in the quality and range of services on offer as technologies improve (eg HTML5, Google’s Accelerated Mobile Pages). Furthermore, greater consumer acceptance of mobile payments and the deregulation of the mobile payments markets, particularly carrier billing, is opening up the market to a wider range of digital goods. Almost half the world’s population has internet access (3.4bn, eMarketer) with approximately three-quarters using a mobile connection, of which almost half is via a smartphone (47.4%), and penetration is forecast to continue to expand at a steady rate. In addition, tablets, wearables and connected TVs are all gaining in popularity.

Analytics group App Annie estimates that in 2016 consumer spend on apps was $62bn and forecasts this to increase to $139bn by 2021. The vast majority relates to games (81% of consumer spend via stores in 2016), although in-app subscription revenues for non-gaming services are forecast to grow even more quickly over the next five years (25% CAGR). Many non-game apps also monetise through advertising or m-commerce, both of which account for a significant share of industry revenues. In 2016, global mobile advertising revenues surpassed desktop to reach $99bn with 27% growth forecast in FY17 (eMarketer) and m-commerce is now almost a trillion dollar market (Euromonitor, 2016).

However, these statistics only relate to the monetisation of apps. In today’s app economy, many businesses distribute apps not for direct monetisation reasons, but to widen their reach or to improve their service offerings. According to Gartner, there will be 269bn app downloads in 2017, up from 102bn in 2013, and developers are adding approximately 400k new apps a year to the two largest app stores, Google Play and the Apple App Store (Statista). AppScatter estimates that there are c 10m apps available to download in app stores, but over 1bn unique URLs from over 6m active publishers; it is already tracking c 90% of the entire market.

The Apple App Store and Google Play are clear market leaders in terms of revenues and downloads. Respectively they have 2.2m and 2.8m apps available (Statista, Exhibits 4 and 5) and, in the US and the UK they account for 85% and 69% of all downloads. However, given the scale of the market, the number of downloads from the ‘other’ category (of c 300 stores) remains significant, particularly outside the US and Europe. China is the world’s largest app market and because most of Google’s services are banned, other platforms account for approximately 75% of all downloads in China (Exhibit 6).

appScatter estimates there are over 300 legitimate app stores worldwide. The lack of independent analytics outside the App Store and Google Play makes it difficult to provide an accurate picture regarding how relevant the other app stores are to developers. However, appScatter’s research, as well as data that we have gathered from other app stores, suggests that as the two giants become over crowded, it is the ‘other’ category that is gaining in relevance. We note below various categories of stores:

The search driven nature of major app stores, which have millions of apps on offer, means that, for all but the leading apps, being discovered can be difficult and expensive.

Alternative stores can improve app discovery as they are less crowded, often drive higher conversions for niche or specialist apps, may have higher daily users in specific geographies (particularly in Asia) and often have less restrictive in-app advertising policies. For instance, a 2013 statistic from the One Platform Foundation suggests that the chances of an app being featured increased by 20-fold on an alternative app store. Similarly, according to the 2013 report ‘Apponomics’,4 data provider Xyo shows that 75% of apps in its search engine get 90% of the downloads, whereas for the major app stores it is more likely to be 10% of apps driving 90% of downloads.

By limiting app distribution to the largest stores, app owners are restricting their reach, but the challenges of managing the wider publication of an app are often seen as too complicated to justify the effort.

We outline what we consider to be the key stages in an app’s journey in Exhibit 10.

For every ‘unique’ app published, there may be tens or hundreds of different versions (instances) created in order to be compatible with different programming languages, operating systems, as well as variations for different geographies (eg for language or compliance reasons) or for adherence to differing app store policies. Publishers will also need to ensure that local legal requirements are adhered to on a store by store basis (such as tax rules).

For instance, a unique app that is published across 10 markets on five different app stores could have 50 versions. For enterprises that publish more than a few apps per year, this can mean managing hundreds of versions of their apps. Developers/publishers will need to register themselves with each app store. Individual approval of each version of an app is required by app stores, a process that can take up to a week. The task of distributing to app stores is generally managed by internal teams and can become resource heavy. Consequently, they focus on stores with the largest market shares, typically just Google Play and the Apple App Store.

Once published, companies need to monitor these apps, not only in terms of performance, but also compliance to a specific company’s policies or branding.

Existing solutions that enable app owners to publish across a number of stores tend to limit services to Android environments where there are many alternative stores to Google Play. For instance, CodeNgo enables distribution to approximately 30 Android stores, and Digital Turbine distributes to eight stores. An open-source initiative was started in 2013 by Opera, SlideME and CodeNgo called The One Platform Foundation for the Android ecosystem although this does not appear to have had much impact. None of these solutions, however, provide feedback on an app’s performance, ranking data or market intelligence.

A further bottleneck relates to app store ranking data and market intelligence, which again tends to be restricted to Apple’s App Store and Google Play.

App Annie’s much cited consumer index shows which apps are most frequently downloaded from App Store and Google Play. Within these stores, data is collected on apps that run the App Annie API, and hence it excludes proprietary apps made by Apple, and some other larger apps such as YouTube. Nevertheless, 90% of the top 100 global app publishers use App Annie, equating to 1,000,000 apps, which provides a deep pool of data. Due to its third-party integrations, App Annie can also offer a range of more sophisticated app analytics (in-app analytics, return on marketing investment), as well as app store optimisation (ASO) tools. However, unlike appScatter, App Annie is not involved in the distribution of an app.

Other analytics players include Yahoo’s free mobile application analytics tool for Google Play and the Apple App Store: Flurry, which is used by 170k developers across 500k+ applications. Flurry offers detailed in-app analytics, although no ranking data as far as we can tell. Other notable offerings include Applyzer, which provides ranking data across iOS, Mac apps, Apple TV and Google Play; Google Analytics (Google Play only); Sensor Tower; Apptopia; Appcodes; MetricsCat; appfigures; AppsFlyer; delta DNA; AppMonsta; and Digital Turbine.

In summary, there does not appear to be any competing solution available that supports both the distribution into and the ranking of apps across as wide a network of stores as appScatter’s.

Through the planned marketplace, appScatter’s licensees will be able to integrate a wide range of applications, which would enable them to consolidate multiple tools into one dashboard.

The appScatter team has worked together in various guises for over a decade, largely led by Philip Marcella, providing consistency and depth of knowledge to the product suite.

appScatter started licensing its platform at the beginning of 2017 and the platform has been revenue generating since January this year, reporting maiden revenues in its H1 results of £0.9m on an operating loss of £2.2m. At the end of June, appScatter was generating annualised recurring revenues (ARR) of £2.1m from c 150 licensees; we have used the platform’s pre-launch experience as a starting point in our forecasts although with limited historical financials, there is a high degree of risk to future forecasts.

We detail our main assumptions below, and given the level of uncertainty surrounding key inputs at this early stage in the company’s development, we also demonstrate what the impact would be on our forecasts should we flex certain key variables such as the conversion rate of registered users to paying licensees and price.

We have built up our forecasts using the following assumptions:

Bar direct marketing and sales commissions, management believes that the large majority of costs should be fixed in nature. Cost of sales (£0.9k in H117) consist of engineering (eight employees), the cost of hosting the platform and mining the data, software for running and securing the platform, and engineering costs related to the ongoing use of the platform.

Operating expenses include staff costs (in addition to engineering headcount there are 16 in sales and marketing, four in admin, and five board members) and the cost of using external services providers related to the development of the platform. The services component within cost of sales is budgeted to drop in FY18, reflecting a lower level of development work being outsourced and the front-loading of work to strengthen the platform prior to public launch. Furthermore, until recently, the company has been incurring the cost of manually publishing apps to those app stores that had not been automated. Developer work to automate the interaction with all 50 supported app stores has now been completed so these costs should also start to come down. We assume that much of this saving will be re-invested in additional employed engineering staff. Sales and marketing costs reflect the £50k per month we assume for direct marketing (CPA) and a similar monthly amount for indirect marketing costs (conference attendance, brand building) and sales commissions (we assume 10%). Total headcount stands at 33; we forecast this to rise to 47 by the end of FY18. To support growth, we have assumed that the cost base will need to grow in FY19 and have factored in a 20% increase in non-staff costs, as well as an increase in headcount to 61.

These assumptions result in very strong revenue growth with gross margins expanding to 65% in FY19 and EBITDA profitability achieved during H119.

The company had no tangible fixed assets on the balance sheet at H117 and expects to spend a small amount on IT equipment for staff (all of which will be expensed) over the whole of FY17. The company does not capitalise development costs; however, it acquired the IP for the appScatter platform from appScatter LLC in 2016 for £0.8m, which is being amortised over three years.

At H117 there was a high receivables balance of £2.9m. This includes £1.1m owing for shares issued in FY16 for services to be provided in FY17 (down from £4.8m at the end of FY16) and a £0.6m prepayment related to IPO fund-raising costs. Trade receivables amounted to £0.6m and £0.1m reflects accrued income (together over two-thirds of H1 revenues). Management expects this to be substantially recovered by the year end and automated invoicing and payment will be introduced with the public launch of the platform, so we have assumed that receivables days normalise to 30 days.

At the end of H117, appScatter had £1.4m of payables and a £0.1m shareholder loan. Part of the £7.4m net IPO proceeds have been used to repay these debts. At 25 September the group reported £6.5m of cash. We forecast that by the year end, assuming the receivables are collected, appScatter will have £3.4m of net cash, sufficient to fund the business to break-even, which we forecast will be reached mid-2019.

The company is in the very early stages of its growth, and it will need to respond flexibly to the market to maximise its chances of building a sustainable business. This means that most of our assumptions are subject to change, in particular the price of the service and the amount it will cost appScatter to attract new licensees. We test these two sensitivities below; however, it is also worth noting that there are a couple of near-term initiatives that we are not yet reflecting in our forecasts:

Indirect sales: appScatter is in discussion with numerous potential strategic partners; however, our forecasts do not factor in any potential growth from partner referrals. We understand from management that it is at an advanced stage of negotiations with a potential strategic sales partner, which has relationships with over 200k app developers. Were the partner to convert even a small percentage of its client base, this would represent a significant step up in licensed users, with no associated CPA.

Marketplace: In our forecasts, we have not factored in any incremental revenues from marketplace add-ins as they have yet to be launched, and we expect a large number to be free (but increasing the stickiness of the platform). However, as explained earlier, commissions earned through referring users to partners could add upside to our revenue forecasts over time.

Data sales: Until pricing points and demand patterns become clearer, we have also excluded from forecasts the potential monetisation of the sale of the raw data being collected. Management expects to start to monetise this data from 2018, and again this may provide another source of upside to our estimates.

Service revenues: The company is generating c £25k per month in services for one customer, related to a range of non-platform related services. We assume this will continue for the remainder of 2017, but have not factored this into our forecasts in FY18. We understand from management that this contract may be extended and so would represent upside to our estimates. We exclude this revenue from ARR calculations.

The conversion rate to date has been higher than the 2% we forecast. It is possible (and management are hopeful) that post launch a higher rate is sustained for longer than we have forecast. All else being equal, were the conversion rate to stabilise at 3%, rather than the 2% forecast, this would imply an operating margin of 31% in FY19 – one of the highest in the sector (Exhibit 12).

appScatter’s pricing has been determined in reference to similar SaaS-based businesses, market testing and in consultation with its client base. We have already factored in an element of price discounting for larger customers in our forecasts. However, in the early years of development, companies often need to adapt their pricing policies as the market becomes established. Using the same CPA and churn assumptions as in our base case forecasts, we have modelled the impact of reducing the monthly rate charged for apps and users by 33% to reflect the possibility that the market is more price sensitive than forecast. We assume a higher conversion rate (3% as above) to reflect the attraction of this lower pricing.

The combination of a higher number of users at a lower price results in a slightly lower revenue profile compared to our base case and the operating margins are clearly affected; in this scenario the group would roughly break-even at the operating level in FY19.

In Exhibit 13 below, we present a selection of SaaS businesses in terms of revenue growth, gross margins, operating margins and sales and marketing expenditure. It is clear that these companies value growth over profitability, as demonstrated by the low operating margins and high levels of sales and marketing expenditure. We therefore think that appScatter may decide to boost its sales and marketing expenditure in order to more rapidly grow its licensee base and gain a dominant position in its niche. It is also with this in mind that we factor in a higher CPA to that experienced by the company to date in our forecasts.

We consider Atlassian to be of particular relevance to our analysis of appScatter, as there are many similarities between the two companies:

appScatter’s valuation hinges on the successful commercial launch of its platform. As this was late in FY17 (and the first month is free), FY17 figures (which we expect in March 2018) will not reflect revenues post this launch. More relevant to track performance at this early stage will be management’s communication regarding the total number of licensees (conversion rates), the exit ARR rate and any updates on the average customer size as well as channel partners.

The shares trade on an FY18 EV/sales rating of 4.1x, a discount to the average of its larger SaaS peer set (6.2x). In FY19, the discount is even more pronounced, trading on 2.0x EV/sales compared to the peer set on 5.1x. While considerably smaller scale than some of these peers, given that this would be appScatter’s first full year of trading we would expect the shares to re-rate upwards towards the peer set on evidence of the successful launch of the platform, and sustained growth during FY18. Our DCF indicates a valuation range of 67p (15% WACC) to 89p (12.5% WACC). At these levels, the shares would still trade at an FY19 EV/sales discount to many of its larger SaaS peers and consequently we do not believe our DCF assumptions to be overly aggressive.

The assumptions underlying our DCF and peer comparisons are outlined in more detail below.

Although approximately 10,000 accounts have registered an interest in using the service, and the company is anticipating a high initial conversion rate after the platform launch, the business is still very early-stage and we consider execution risk to be relatively high. Forecasts are highly sensitive to fairly small changes in conversion rate, pricing and CPA. In our valuation of the group, we have tried to capture forecasting uncertainties by using a higher CPA than that experienced to date by management as well as by using a relatively high (15%) discount rate in our DCF, which returns a value of 67p. With the demonstrable success of the platform’s commercial launch and as the group moves towards break-even, we would expect the rate the market applies to come down fairly rapidly and so we also demonstrate the impact on the DCF valuation should we vary the discount rate. For example, using a 12.5% WACC (which we consider more typical for technology companies where the demand has been established but which are still loss making) we derive a valuation of 89p per share.

Exhibit 14 shows a DCF matrix defined by discount rates and by terminal EBITDA margins. We model a 10-year period, with revenue growth fading steadily post the explicit three-year forecast period to 2% in perpetuity. Even though we use a fairly long 10-year forecast period, c 53% of the enterprise value for our base case valuation lies in the terminal period.

Key DCF assumptions:

The DCF is sensitive to small changes in key variables. For instance, the same methodology applied to the ‘higher conversion rate’ scenario outlined earlier returns a value of 107p per share (15% WACC) or 138p share (12.5% WACC). The lower pricing scenario returns 51p (15%) and 70p (12.5%).

In the absence of direct competition, we have selected a range of peers, which we group into three categories. Many companies could feature in multiple categories.

appScatter trades on an FY18 EV/sales multiple of 4.1x, a discount to the larger (more mature) SaaS peer set (6.2x). Assuming the group progresses in line with forecasts, this discount expands in FY19 when it trades on 2.0x sales (peers 5.1x) and 15.1x EV/EBITDA (peers 35.1x).

At our DCF valuation of 67p, the implied multiple of 5.7x FY18e EV/sales for appScatter is more in line with its peer average, although still at a discount to more established SaaS companies, notably Atlassian, Xero and Workday, which trade in the 6.0x to 10.4x range, which we feel is justified due to the early stage, unproven nature of the appScatter platform. However, if the company delivers in the short to mid-term, an upwards re-rating towards these multiples may be justified. At our 89p DCF valuation (12.5% WACC), this would imply an FY18 EV/sales multiple of 7.7x.

appScatter is a development-stage business, with a scalable business model moving into commercialisation in a potentially very large market. Consequently, there are significant risks, uncertainties and potential. Our forecasts and the valuation of the company will be sensitive to the following factors:

We have also discussed in the financial section the sensitivity of our forecasts in relation to the pace of adoption, the cost to acquire customers, the rate of conversion to paying customers, renewal rates and the price of the service. While we have not explicitly modelled it, there is the risk that the pace of adoption is significantly below our base case, in which case the company would struggle to reach profitability.