Summary

  • 01 SYSTEM SRL is a French managed-IT company founded in 2008, based in Miribel near Lyon. Its registered activity is computer-facilities management, while its current offer combines managed support, business connectivity, voice, security, equipment and hosting. That combination matters more than any single telecom label.
  • The company has real network-operating evidence. AS60502 dates from 2017, currently originates a block equivalent to 1,024 IPv4 addresses, has two visible upstreams, and is listed at France-IX AURA. This gives 01 SYSTEM more routing control than a pure reseller, but it does not eliminate dependence on outside carriers or prove physically diverse fibre paths.
  • 01 SYSTEM says its Miribel data centre has 28 live racks, room for 84, redundant power and cooling, multi-operator connectivity and a 99.95% availability target. Those claims support a credible regional continuity proposition. They remain management claims, however, and some public figures vary by page; buyers need scope, measurement and test results rather than labels alone.
  • Revenue is not publicly disclosed in the available summaries, so rapid top-line growth cannot be established. Filed-account summaries show net profit rising from EUR232,000 in 2024 to EUR423,000 in 2025, cash of EUR509,000 and financial debt down to EUR186,000. The balance sheet looks stronger, but product margins, recurring revenue, concentration, utilisation and maintenance capital remain unknown.
  • The central test is whether customers willingly pay a premium for an accountable local bundle. 01 SYSTEM can create value when one contract genuinely reduces outage duration, supplier coordination and staffing needs. It destroys value if long commitments, vague service levels and liability limits leave the customer carrying most of the loss while the provider charges more than credible fibre, cloud and specialist alternatives.

Independence has a price

The easiest way to sell business internet is also the least defensible. Buy a wholesale line from a larger carrier, add a support number, mark up the monthly charge and hope the customer never asks who controls the repair. That model can generate revenue with little capital, but it gives the reseller few levers during an outage. The access owner controls the physical intervention. The transit provider controls the wider route. The equipment vendor controls the replacement. The reseller owns the apology.

01 SYSTEM has chosen a more ambitious position. Its public materials describe a managed-service company with its own data centre, autonomous system, Internet number resources, support operation and ability to combine connectivity with hosting, backup, security, voice and equipment. The economic purpose of that stack is not technological prestige. It is to bring more failure points within one commercial and operating boundary, then charge customers for the reduction in coordination time.

That can be valuable in the small and midsize market around Lyon. A manufacturer, accountancy practice, regional distributor or professional-services firm may have no network engineer of its own. When a site loses access to its applications, the customer does not want a debate among the fibre carrier, firewall supplier, hosting company and software contractor. It wants one party to diagnose the whole path. A provider able to see the local network, the wide-area connection, the hosted server and the backup can often resolve the fault faster than four suppliers acting separately.

But integration moves risk as well as convenience. The customer pays 01 SYSTEM; 01 SYSTEM pays carriers, electricity suppliers, equipment vendors, software providers, landlords or facility contractors, and its own staff. The company keeps the spread after support and infrastructure costs. If it has enough customers on standardised platforms, it can share engineers and spare capacity across them. The customer benefits from avoided recruitment, reduced capital spending and shorter outages.

The downside is asymmetrical. A service provider can limit its contractual damages to a few months of fees. The customer can lose a day's orders, miss a production run or fail to serve its own clients. A second fibre route benefits the customer when the first fails, but the provider carries the standing cost every day. Spare switches, batteries, generator tests and backup capacity are valuable precisely because they are usually idle. The commercial challenge is therefore to make customers pay for absence: the incident that did not become a crisis.

That is a difficult sale. France's business-access market contains cheap mass-market fibre, dedicated products with repair guarantees, larger national integrators and public cloud priced by the hour. A regional provider cannot rely on the word "local" alone. It must show that local control changes the probability, duration or consequence of failure. Strategy without that proof is simply a more expensive bundle.

What the company is, and where its responsibility ends

The legal identity is clear. France's national company directory records 01 SYSTEM under SIREN 504 228 271, with its active head office in Miribel and the activity code 62.03Z, management of computer facilities. It records incorporation in 2008 and a workforce band of 50 to 99 employees for 2023. The company's legal notice gives the same registration number, a Miribel address, EUR500,000 of capital and the commercial name ONE SYSTEM. The entity was originally a limited-liability company and now presents as a simplified joint-stock company with one shareholder.

Its registered purpose is broader than network access. Public corporate data describes audit and advice, training, change support, engineering projects, managed services and maintenance of customers' information systems. This makes "regional ISP" a useful economic category but an incomplete description. 01 SYSTEM appears to be an IT operator that uses connectivity and hosting as anchors for a wider managed relationship.

The current service catalogue supports that reading. It lists managed support and supervision, fibre, ADSL and mobile backup, Wi-Fi, internet voice, hardware, common workplace software, secure hosting and domain-name management. The infrastructure page adds maintenance, cyber services, hosting, connectivity, voice and supervision. The company seeks a tailored quote rather than publishing a simple price list, which is consistent with project work and recurring service contracts rather than a self-service access business.

The operating boundary is still important. 01 SYSTEM can control its engineers, routers, hosted equipment and service desk. It does not control every trench, electricity substation, semiconductor lead time, software patch or upstream route. Its own standard terms allow subcontracting. Its data centre page calls the facility carrier-neutral, which by definition means customers and the operator rely on external networks entering the site. Even when a provider owns the customer relationship, many critical inputs remain purchased.

Nor should every brand claim be assigned automatically to the legal entity without context. The company says on its website that it has 800 customers, 750 projects a year, 420 terabytes of protected data and 98% satisfaction. Its LinkedIn profile refers to 60 employees, while the platform displays 44 employee profiles and the government register gives a historical band of 50 to 99. These numbers may use different dates and definitions. They are useful indicators of intended scale, not an audited operating schedule.

There is a similar variation in facility descriptions. One current page refers to 200 square metres over two levels; the home page and social profile have used 184 square metres. The detailed page says 28 racks are in service with capacity for 84, while headline copy can make 84 sound like the facility's present size. None of these differences makes the facility unreal. They show why a customer should contract against the deployed configuration it is buying, not against a rounded marketing statistic.

A bundle designed to make one customer worth more

The company's business model has a sensible progression. A fibre circuit opens the account. Managed support adds a recurring fee. Firewalls, Wi-Fi and voice increase the number of managed devices. Hosting moves workloads into 01 SYSTEM's facility. Backup and recovery add storage and monitoring. Equipment refresh and software renewals create periodic sales. Security services attach further recurring work. Each layer raises revenue per customer and makes the relationship harder to displace.

This is not necessarily lock-in in the abusive sense. Integration can remove real cost. An external support team that already knows a customer's switches, identity setup, servers and backup policy should diagnose incidents faster than a new contractor. A provider can spread specialised engineers across hundreds of clients that could not each hire them. Standard monitoring and purchasing can reduce unit cost. The data centre can pool redundant power, cooling and security more efficiently than each small business can do on its premises.

The model becomes less attractive when every additional product deepens dependence without improving outcomes. A customer that buys access, voice, security, hosting, backup and support from one company has concentrated its operational risk. If the provider's service desk, billing dispute or security controls fail, several services can fail together. The one-supplier bundle therefore needs stronger transparency and exit provisions than a collection of replaceable products.

01 SYSTEM's publicly posted managed-service conditions illustrate how the economics can work. They describe monthly, quarterly or annual recurring fees. User support is priced according to user count and equipment complexity; administrator support depends on server and workstation counts, network and security complexity, and the chosen scope; executive reporting depends on frequency. The company can also provide staff by the day or quote separately for projects outside the recurring fee.

This is rational cost allocation. More users generate more tickets. More servers and security products create more monitoring and patching. Complex sites consume scarce senior-engineer time. A fixed recurring fee gives the provider predictable cash and gives the customer a budget. Separate quotes prevent unusual projects from consuming the support pool.

Yet the customer should ask what the fee buys at the margin. Does adding a server always add the same support burden? Are automated monitoring and standard patching priced as if they require manual labour? Does the provider earn extra margin when incidents fall, or must it demonstrate prevention to renew? A contract tied only to device counts can reward infrastructure sprawl. A better arrangement combines a base fee with clear availability, response, recovery and project outcomes.

Hosting adds another layer of recurring economics. The posted hosted-services terms say service begins after an accepted quote and signed contract; rent and services are specified in the quote, with bandwidth added where necessary. The version online provides for payment in advance by quarter, an initial 60-month term, automatic renewal and repricing by reference to the Syntec index and configuration. These are older public terms and may not govern every current customer. They nevertheless show the commercial logic: long contracts help recover setup and infrastructure costs, while advance billing supports working capital.

The same terms can burden the buyer. A five-year commitment is long in a market where hardware prices, fibre availability and cloud products change quickly. A request to alter the hosted service was described as restarting the 60-month term. That protects the provider's investment but can make normal growth expensive to unwind. A well-advised customer will negotiate portability, data-export format, migration assistance, deletion evidence, price review, termination for repeated service failure and treatment of equipment or licences at exit.

The network is evidence of capability, not a complete identity

The strongest independent evidence of 01 SYSTEM's network role is AS60502. Public registration records associate the autonomous system with 01 SYSTEM SRL and an organisation record created in 2017. The RIPE record classifies the company as a Local Internet Registry. That means it participates directly in the regional framework for Internet number resources rather than borrowing every address and route from a retail carrier.

Current route summaries show a block equivalent to 1,024 IPv4 addresses originated by the network. The observed routes include the covering 185.221.224.0/22 and two more-specific /23 announcements. Those announcements should not be added together as if the company has 2,048 addresses: the /23 routes are subdivisions of the /22. This is a modest footprint compared with a national carrier, but enough to support hosting, infrastructure management and customer services.

The RIPE route policy and current BGP observations identify LASOTEL and ADELI as upstreams. The record also contains an import from AS43100, the autonomous system associated with the regional exchange environment. PeeringDB lists 01 SYSTEM at France-IX AURA and at two regional facilities, LASOTEL PIXEL in Villeurbanne and Maxnod in Saint-Trivier-sur-Moignans. Its profile describes a regional network, a 1-5 Gbps traffic band and an open peering policy, though several profile fields were last updated in 2021 or 2022.

This topology is economically meaningful. Two upstream relationships can protect against a commercial or routing failure at one carrier. An exchange connection can keep some local traffic off paid transit, reduce route length and improve access to networks present in the same peering community. A provider with its own address space can change transit suppliers without renumbering every hosted service. It can set route preferences, announce more-specific routes during an incident and maintain a recognisable abuse and network contact.

None of that proves end-to-end redundancy. Two upstream companies may enter the building through the same duct, use common long-haul infrastructure or depend on the same exchange. A PeeringDB facility listing does not prove that every customer product is dual-homed at both sites. A route-policy statement records intended relationships; live performance can change. An autonomous system is control over routing policy, not ownership of fibre all the way to the customer.

The public record also raises an IPv6 question. PeeringDB lists IPv6 capability and an IPv6 address at France-IX AURA, but current route summaries viewed for this research did not show an IPv6 prefix originated by AS60502. ARCEP's 2025 hosting-provider IPv6 dataset, based on December 2024 samples, attributed 0% IPv6 accessibility to the 30 sampled websites and seven sampled mail domains hosted on the network, and 0% to ten sampled DNS domains. The samples are small and historical; they do not measure every service. They do suggest that IPv6 enablement at an exchange or in a profile had not translated into broad hosted-service availability.

That gap matters commercially, though not because IPv6 immediately determines revenue. It is a test of execution. A provider that sells technical stewardship should be able to show a roadmap for dual-stack hosting, monitoring, security policy and customer migration. Lack of public IPv6 origination can also make future address growth more dependent on scarce IPv4 resources. The fact would change the judgment positively if 01 SYSTEM began originating stable IPv6 space and demonstrated material customer adoption.

Resource ownership itself is not the expensive part. The RIPE NCC charging scheme for 2026 sets EUR1,800 per Local Internet Registry account, plus defined resource charges. That fee is trivial beside payroll, carrier contracts, power or equipment. The cost lies in operating the capability: competent engineers, secure route configuration, 24-hour monitoring, incident handling, abuse response, registry maintenance and spare network capacity.

Reliability is a cost stack

01 SYSTEM's data centre page gives the physical case for its premium. The company says the Miribel site covers 200 square metres on two levels, has 28 active 19-inch racks and can expand to 84. It describes two independent power circuits with uninterruptible supplies and generators, N+1 cooling with hot- and cold-aisle containment, free cooling, multi-operator connectivity, biometric and badge access, video surveillance, guarding, early smoke detection and inert-gas fire suppression.

Each item addresses a different failure mode. Dual power paths help only if they are genuinely independent through switchgear, batteries and generator supply. N+1 cooling means one cooling component can fail without exhausting capacity, but only within the design load. Fire detection and gas suppression can limit physical damage. Controlled access reduces human risk. Carrier neutrality gives a customer more network options. None is a substitute for the others.

The company also advertises 99.95% availability, recovery-time objectives of no more than one hour, recovery-point objectives of no more than 15 minutes, and failover involving sites in Venissieux and Maxnod. It says major failover can occur in under 30 minutes. Those are strong claims for a regional provider. At continuous measurement, 99.95% allows about 4 hours and 23 minutes of downtime in a 365-day year. An RPO of 15 minutes describes possible data loss, while an RTO of one hour describes restoration time; both depend on the service actually purchased, replicated and tested.

The public contract record shows why scope matters. The 2019 hosted-services terms available on the company site state a 98% availability commitment during the technical-support window, equivalent in the document to 7.3 days a year, and characterise the obligation as one of reasonable means. The newer data centre page advertises 99.95%. These figures may apply to different products, periods or contract generations. A buyer should not assume the website target replaces signed terms. It should ask for the current service schedule, excluded maintenance, measurement point, credit formula and recovery test evidence.

Redundancy also consumes capital before it generates revenue. The operator needs switchgear, uninterruptible-power systems, batteries, generators, fuel arrangements, cooling plant, fire systems, racks, cabling, routers, security appliances and monitoring. Much of it has a finite life. Batteries degrade. Servers and storage age. Vendor support expires. More capacity must be installed before racks are full, yet unused capacity earns nothing.

The company's claimed path from 28 racks to 84 is therefore an option, not automatically an asset. If demand arrives, expansion can spread fixed facility costs over more customers and improve return on invested capital. If demand does not arrive, the building, power envelope and cooling preparation remain underused. If management fills racks with low-margin commodity hosting, utilisation rises without necessarily creating much value. The crucial metric is contribution after power, transit, hardware depreciation and support, not rack count.

Energy is a material variable. The company says its data centre achieved a power usage effectiveness ratio of 1.14 in 2024 and has 90 kWp of solar panels. PUE divides total facility energy by IT-equipment energy; lower is better. An ADEME study of French cooling technologies modelled PUE ranges around 1.36-1.39 for air-cooled chillers and 1.24-1.27 for water-cooled systems with free chilling. That does not disprove 01 SYSTEM's figure: facility size, load, climate, measurement boundary and technology differ. It makes the 1.14 claim valuable enough to verify with a full-year energy series and a clear boundary.

Electricity prices can move the margin quickly because power is both a direct input and the source of cooling load. Eurostat reported that EU non-household electricity prices for medium consumption users averaged EUR18.37 per 100 kWh in the second half of 2025, while France recorded a 14.1% year-on-year decline. A regional data centre's contracted tariff and consumption band may be very different from that benchmark. The broader point is that lower energy cost can temporarily lift profit, while a reversal can squeeze fixed-price customer contracts.

Regulation is adding another fixed cost. The European Commission says data centres with installed IT power demand above 500 kW are subject to public reporting requirements under the Energy Efficiency Directive. 01 SYSTEM has not publicly stated an installed IT load in the sources reviewed, so applicability cannot be assumed. Even below that threshold, customers and regulators increasingly expect energy, water and efficiency evidence. Measurement is becoming part of the product.

The accounts show improvement, but not the engine

Revenue is the largest missing number. Public summaries of the filed accounts say 01 SYSTEM used partial confidentiality, and they do not disclose annual sales. Pappers explains that the 2025 filing met at least two of the applicable thresholds for balance sheet, net sales and employees. The result cannot be used to infer a precise revenue figure or current headcount. It means only that a full public income statement is not available through that summary.

The disclosed financial series is still informative. Pappers reports net profit of EUR148,000 in 2022, EUR244,000 in 2023, EUR232,000 in 2024 and EUR423,000 in 2025. Profit therefore dipped about 5% in 2024 and rose about 82% in 2025. Cash was EUR509,000 in 2025, up from EUR458,000 in 2024 but below EUR558,000 in 2023. Financial debt fell from EUR1.21 million in 2022 to EUR1 million in 2023, EUR496,000 in 2024 and EUR186,000 in 2025.

That is a meaningful strengthening. The company appears to have generated profit while reducing financial leverage. A net cash position before other liabilities gives it more flexibility to replace equipment or absorb an incident. Working capital was EUR776,000 in 2025, with a reported coverage ratio of 1.7. This does not look like a provider relying on ever-increasing bank debt to keep old equipment running.

It is not enough to establish value creation. Without revenue, profit margin cannot be calculated. Without cash flow, the conversion of accounting profit into cash is unclear. Without a fixed-asset note, one cannot separate data-centre plant, leased equipment and ordinary office assets or see depreciation by class. Without segment data, one cannot know whether profit came from managed support, equipment resale, hosting, connectivity, a one-off project, lower power prices or reduced financing cost.

The balance sheet offers clues. A separate public summary showed about EUR1 million of net tangible fixed assets across recent years and total assets of roughly EUR3.5 million at September 2024. Supplier payables rose to about EUR634,000 in 2024 from EUR520,000 in 2023, while customer receivables rose to roughly EUR929,000 from EUR818,000. Those are rounded secondary figures, not a substitute for the filed notes. They indicate that hardware and project activity can tie up cash even when debt falls.

The economic question is whether the company earns more than the cost of maintaining its asset and service promise. A data centre can report profit while deferring battery, generator, server or network refresh. Conversely, a year of heavy investment can depress cash while improving future resilience. The missing measure is maintenance capital: the annual spending required just to preserve present capacity and service quality. Growth capital for racks beyond the current 28 should be considered separately.

Revenue growth must also be separated from customer value. The company says its workforce and sales have doubled since 2016, but provides no annual series. Even if accurate, nominal doubling over a decade does not reveal organic volume, inflation, acquisitions or mix. A higher invoice can reflect more services, higher supplier cost or more lock-in. It creates value when the customer receives lower downtime, stronger recovery, reduced staffing cost or better security worth more than the increase.

The 2025 profit step is encouraging because reliability is easier to fund from earnings than from hope. It becomes convincing if accompanied by retained customers, rising recurring revenue, healthy service margins, on-time refresh and stable incident performance. If it came mainly from delaying investment or one-off projects, it says less about the durability of the model.

What customers are really paying for

ARCEP's 2024 market results provide a useful baseline. The regulator reported 2.1 million business fixed-internet subscriptions, with fibre representing 84% of business access at the end of 2024. An enterprise paid an average EUR93 before tax per month for standard high- or very-high-speed internet, and EUR315 for a more elaborate offer with added features, quality or multi-site capability.

Those averages are not a quotation for 01 SYSTEM and do not include a full managed stack. They show the pricing problem. Basic access is cheap enough that a regional provider cannot finance field support, dual transit, hosting and senior engineers from a small connectivity markup. The provider must move customers toward higher-value contracts, or accept a commodity margin.

The premium can be rational. ARCEP describes dedicated business fibre as providing guaranteed bandwidth and a controlled repair time, usually four hours. Orange Business advertises dedicated symmetric fibre with a four-business-hour restoration guarantee. A customer can therefore buy accountable connectivity from a national operator without buying local hosting or managed IT. 01 SYSTEM has to outperform that substitute through faster diagnosis across the whole estate, local intervention, flexibility or a better total price.

The customer should compare total operating cost, not line items. One option is basic fibre plus an in-house administrator. Another is dedicated fibre plus a national support contract. A third is 01 SYSTEM's combined access, hosting and support. A fourth is public or dedicated cloud plus an independent local managed-service provider. A fifth is customer-owned hardware in a neutral colocation facility.

Each alternative allocates cost differently. An in-house administrator is fixed payroll but can understand the business deeply. A national carrier has reach and procurement scale but may divide responsibilities across teams. A public cloud offers rapid capacity and transparent unit prices but can require scarce architecture skills and expose the customer to egress and support charges. Colocation gives control but leaves hardware ownership and on-call coverage with the customer. The 01 SYSTEM bundle is attractive when it removes enough coordination and staffing cost to offset its likely premium.

Large providers set a visible commodity floor. OVHcloud advertises France-hosted dedicated servers starting below EUR100 before tax per month for entry configurations, while enterprise products offer a 99.95% service level and included anti-DDoS protection. This is not equivalent to an on-site engineer, managed firewall or regional recovery plan. It limits how much of 01 SYSTEM's hosting price can be justified by hardware and bandwidth alone.

The same discipline applies to reliability. Customers should put a price on an hour of outage, then identify which part of that loss the provider can actually prevent. If an office of 40 people costs EUR2,000 an hour in idle payroll and lost contribution, paying several hundred euros a month for genuine path diversity and tested recovery may be sensible. If the second circuit shares the same duct, the backup has never been restored, or support responds without resolving, the premium buys appearance rather than resilience.

Contracts decide who carries the outage

Marketing says who benefits; contracts say who pays when the benefit fails. 01 SYSTEM's current general terms published in 2023 describe its obligations as obligations of means rather than guaranteed results. They limit potential compensation to fees for the affected services over the preceding six months and exclude indirect economic, financial or commercial losses. They also put insurance obligations on the customer for theft, fire, electrical risk, cyber incidents and data reconstruction.

These provisions are common in IT services because the provider's monthly fee is small relative to a customer's possible business loss. Unlimited liability would make many services uneconomic or uninsurable. The clauses still mean that the customer retains most outage risk. A six-month fee refund will not compensate a factory for missed output or a retailer for lost sales.

The provider's older hosted-services terms go further by excluding responsibility for failures of internet transport operators and access suppliers, and by describing bandwidth fluctuation as outside its technical control. That is precisely where the value of owning network capability must be tested. An autonomous system and multiple upstreams should allow 01 SYSTEM to mitigate some carrier failures. If the contract excludes all meaningful external faults, the customer should ask what measurable risk the connectivity premium transfers.

The fair answer is not to demand that 01 SYSTEM insure every business consequence. It is to align price, control and remedy. The company should be responsible for architecture and operations it controls, transparent about shared risks, and required to provide service credits or termination rights after repeated failure. The customer should insure its own consequential loss, maintain tested continuity plans and avoid placing every critical service in one failure domain.

Data obligations add another layer. The CNIL's guidance on subcontractors says customers should require sufficient security guarantees, clear contracts, incident notification, audit rights, encryption and attention to the full subcontracting chain. Its backup guidance recommends regular tested copies, at least one geographically separate copy, at least one offline copy and secondary-site replication where availability requirements are high.

Those recommendations make local hosting valuable only when the architecture is demonstrably separated. A backup in another rack in the same room is not geographic resilience. A second site reached through the same credentials can still be exposed to one compromise. A customer should know where its primary copy, replicated copy and offline copy reside; who can delete them; how often restoration is tested; and whether the recovery site depends on the same upstream, electricity region or management platform.

Suppliers make independence relative

01 SYSTEM's public partner material names 3CX, HP, Microsoft, Veeam, HPE and Stormshield, while its data centre material names Darktrace. These relationships can improve purchasing, expertise and support. They also expose the company to vendor pricing, licence changes, product end-of-life decisions and channel conditions. A regional provider does not become independent by purchasing more products; it becomes more capable of managing the dependencies.

The network has the same structure. LASOTEL and ADELI appear as current upstreams. France-IX AURA adds peering. These are sensible layers, but their independence must be mapped physically and commercially. The most useful diligence document would show carrier, handoff, building entrance, duct route, upstream autonomous system, exchange connection, router, power feed and failover policy for each critical service.

Supplier concentration can affect margin before it affects availability. A software vendor can increase per-user fees; a hardware vendor can end support; a carrier can reprice a circuit; an electricity contract can reset. If 01 SYSTEM has fixed customer prices for several years, it carries the squeeze. If contracts pass every increase through, customers carry it. Syntec-linked annual repricing protects labour margin but does not necessarily match equipment, energy or software inflation.

The decline in financial debt improves resilience, but the absolute cash balance is not large relative to a major facility refresh or prolonged incident. EUR509,000 of cash can fund many ordinary replacements; it can disappear quickly in a generator, battery, storage or server programme. The company may lease equipment, use vendor finance or draw on its shareholder. Public data does not show committed facilities or supplier credit terms.

Human capital may be the tightest supplier of all. Managed service requires broad skills across endpoints, networks, security, cloud, backup and customer applications. The company publicly recruits systems administrators and emphasises training. Recruitment is a positive growth signal and a reminder that senior engineers are costly. If service revenue grows faster than skilled staffing and automation, response quality can fall before reported profit does.

Customers, concentration and the regional ceiling

The company says it has more than 800 customers in Auvergne-Rhone-Alpes. If the claim represents active paying organisations, it suggests a broader base than a provider dependent on three large accounts. It does not reveal concentration. One hosting customer or large managed-services contract can contribute more than hundreds of small equipment buyers. Project count is equally ambiguous: 750 annual projects may range from minor installations to multi-year transformations.

Geographic concentration is both the advantage and the ceiling. Miribel is close enough to Lyon and the Ain industrial base for local intervention. Proximity supports trust, referrals and field service. It also concentrates sales exposure in one regional economy and puts physical infrastructure near the same weather, grid and transport environment as many customers.

The company can expand remotely across France for services that need little field work, but its differentiation weakens with distance. A customer in Lille or Bordeaux can buy from a local specialist, national integrator or cloud provider. To grow beyond its home region, 01 SYSTEM must either develop repeatable remote products, build a partner network or open more field capacity. Each path changes the economics.

Customer mix matters as much as geography. Small firms value outsourced competence but can be price-sensitive and financially fragile. Mid-market industrial customers pay more for continuity but demand formal service levels, audits, recovery tests and procurement evidence. Public-sector or health workloads can add certifications and tender cost. The provider's catalogue spans all of these possibilities, but public sources do not disclose revenue by sector or contract type.

One low-weight market signal comes from a 2026 LinkedIn post by a regional security executive, who said his company's infrastructure had been installed in 01 SYSTEM's data centre and thanked its teams. That is evidence that at least one external business publicly associated a real installation with the facility. It is not proof of broad satisfaction, uptime or concentration.

The company's own social profile is another signal, not verification. It says sales and staff have doubled since 2016, highlights active hiring and presents local hosting as an alternative to large foreign platforms. It also makes sweeping claims about having no dependence on major US technology companies while the service catalogue and partner pages openly include Microsoft and HP products. The tension does not invalidate the local-hosting proposition. It shows that sovereignty should be defined contractually as data location, legal control, operational access and replaceable dependencies, not as an absolute slogan.

The absence of a detailed public customer roster is reasonable for a security and infrastructure provider. The absence of anonymised retention, concentration, incident and recovery statistics is more consequential for economic analysis. A business selling trust can disclose aggregate evidence without exposing customers. Renewal rates, top-ten revenue share, recurring-revenue proportion, support response percentiles and successful restore tests would make the premium easier to evaluate.

Regulation creates both work and demand

Compliance is a cost centre and a sales opportunity. As a host and managed-service provider, 01 SYSTEM must maintain contracts, security controls, incident processes, data-protection obligations and telecommunications administration. Customers facing the same burden may pay the company to handle part of it.

The RGPD framework is especially relevant because IT providers often process or can access personal data for clients. The CNIL expects the customer to know a provider's security measures and subcontracting chain. That favours a local operator willing to permit audits and explain where data resides. It also raises the operator's cost: documentation, access control, logging, incident notification, deletion and return processes must be maintained across staff and subcontractors.

NIS2 and France's emerging cyber rules may expand demand for mapping, continuity and supplier oversight. Whether a particular obligation applies depends on service type, size and legal scope; the public evidence reviewed here is not enough to classify 01 SYSTEM conclusively. The commercial direction is clear. Regulated customers will ask more questions about provider dependencies, incident reporting and recovery. Providers that can answer will win work; those that rely on broad assurances will face longer sales cycles or exclusion.

Network regulation is also changing the market. ARCEP's framework for 2024-2028 aims to manage copper retirement, complete fibre deployment and increase competition in business services. The regulator says competition in enterprise fibre remains insufficient and requires offers with enhanced quality across fibre networks. This creates wholesale options for regional providers, but it also gives customers and rival integrators more alternatives.

Copper retirement can generate migration projects and new recurring fibre revenue. It can also burden the support desk as old voice, alarm, payment and machine connections move to IP. A provider with local field teams can monetise the transition. The work creates value only when customers receive a stable replacement, not when migration is treated as a one-time equipment sale.

Environmental reporting may similarly become a differentiator. ADEME estimates that data centres in France consumed about 10 TWh a year in its 2026 outlook and expects demand to rise substantially. Customers will increasingly compare energy efficiency, renewable supply, equipment life and heat use. 01 SYSTEM's solar and PUE claims could support the sales case if independently measured. If not, they remain exposed to better-documented competitors.

Competition comes from four directions

The first competitors are national telecom operators. They can bundle dedicated fibre, mobile backup, voice, security and service guarantees across many sites. Their purchasing and network scale are difficult to match. 01 SYSTEM's defence is local accountability and the ability to manage the customer environment beyond the access line.

The second are national cloud and hosting providers. They offer visible prices, large capacity, multiple facilities and broad product catalogues. Their weakness for an SME can be the responsibility gap: they operate the platform but may not manage the customer's office, applications or users. 01 SYSTEM can sit above them as a manager, compete with its own hosting, or use both in a hybrid design.

The third are regional integrators and managed-service firms. They may have similar customer relationships without owning a data centre or autonomous system. Their asset-light model can be cheaper and more flexible. 01 SYSTEM must earn a return on its facility by proving that direct operating control improves outcomes.

The fourth substitute is the customer itself. A sufficiently large business can hire administrators, contract directly with carriers, place equipment in colocation and maintain cloud accounts. This offers control and avoids provider margin, but it requires recruitment, coverage and governance. For many smaller customers, the in-house option costs more than a shared service. For larger customers, it can be the strongest negotiating alternative.

Competition therefore prevents a simple "own more, earn more" conclusion. Owning a data centre and network adds fixed cost. The asset creates an advantage only when enough customers pay for it and utilisation does not compromise redundancy. An asset-light rival can undercut commodity services; a larger rival can outspend on scale. The regional operator survives in the middle by combining intimacy with credible engineering.

The facts that would change the judgment

The present judgment is cautiously positive. 01 SYSTEM has more substance than a branded reseller: a long-lived French company, an identifiable regional facility, its own routing presence, multiple visible network relationships, recurring service contracts, improving profit and declining financial debt. These are foundations for a valuable regional continuity business.

The evidence does not establish that customers pay enough for the full cost of reliability. Revenue and segment margins are unavailable. Facility claims are not accompanied by a current independent assurance report. Public service-level language varies. Customer concentration, churn, utilisation, incident history, capital expenditure and physical route diversity are unknown. The company may be earning an attractive return on a disciplined integrated stack, or it may be cross-subsidising expensive infrastructure with project and equipment sales.

Several disclosures would strengthen the case materially:

  1. Recurring revenue growing faster than project revenue, with stable or rising gross margin after carrier, power, software and hardware costs.
  2. Customer retention above 90%, low top-ten concentration and evidence that clients add services voluntarily rather than remaining because migration is difficult.
  3. A current service schedule showing 99.95% or better availability for the relevant hosting product, with measurement rules, actual performance and meaningful remedies.
  4. Successful annual recovery tests demonstrating the advertised recovery objectives across genuinely separate sites and credentials.
  5. A physical-path map confirming that upstream carriers, building entrances, ducts, routers and power feeds do not share hidden single points of failure.
  6. Stable IPv6 origination and growing hosted-service adoption without weakening security or support.
  7. A maintenance-capital plan showing timely replacement of batteries, cooling, generators, servers, storage and network equipment while preserving net cash.
  8. Independently measured full-year PUE and energy data using a stated boundary, together with utilisation and renewable-energy evidence.
  9. Support statistics showing response and resolution percentiles, not just average satisfaction, alongside engineer retention and on-call coverage.
  10. Clear data portability, subcontractor disclosure and exit assistance that let customers leave without operational hostage costs.

The negative triggers are the mirror image: repeated outages without transparent post-incident reports, a rise in supplier payables or debt while refresh is deferred, loss of an upstream without replacement, dependence on one large customer, persistent absence of IPv6 planning, or contracts that promise less than the website suggests. Any of these would make the ownership strategy look like overhead rather than resilience.

The deepest economic point is that reliability is not a product sitting in a rack. It is an allocation of resources before failure and responsibility after it. 01 SYSTEM's network and facility give it tools to reduce outage risk. Its local staff can reduce coordination delay. Its improving balance sheet can support reinvestment. Customers should pay for those advantages when they are measured and contractually attached to the service.

They should not pay merely because the provider owns an autonomous system, a building or a regional story. Resource records prove operating capability, not customer value. Racks prove capacity, not uptime. Redundant diagrams prove design, not physical separation. Profit proves that the company kept money, not that customers received more than they paid.

01 SYSTEM's opportunity is to make local accountability auditable. If it can show that its integrated model restores service faster, protects data better and costs less than assembling the same result from separate suppliers, the capital burden becomes a moat. If it cannot, larger carriers will sell guaranteed access, cloud companies will sell cheaper capacity and asset-light integrators will sell the same relationship without carrying a data centre. Owning reliability is worth the price only when customers can see exactly what they are buying.