Document type: Toolkit Practice area: Technology — Platform Regulation Jurisdiction: European Union, with United States comparison Last reviewed: 5 September 2026
How to use this toolkit
European platform regulation is unusually form-driven. What a compliant notice must contain, what a statement of reasons must say, what a transparency report must include — these are prescribed, and a company either produces the required elements or does not.
So this toolkit is mostly forms and specifications. The annotations explain what each element is doing, because a team that understands why a field exists fills it in properly, and a team that does not produces compliant-looking documents that fail on inspection.
Tool 1 — Scoping memo template
Write this first. It settles the internal argument that otherwise consumes six weeks.
EU PLATFORM REGULATION — SCOPE DETERMINATION
[Company] — [Date] — Prepared by [ ] — Privileged
1. SERVICES OFFERED
Service | Description | EU recipients? | Est. share
-----------------|--------------------|----------------|------------
[each service, separately — a company often has more than one tier]
2. TERRITORIAL SCOPE
Do we offer services to recipients located in the Union?
YES / NO — basis: __________________________
NOTE: scope follows the OFFERING, not establishment. Absence of
an EU entity, staff, or revenue is NOT a scope answer.
Indicators considered: EU-language interface, EU currency, EU
payment methods, EU targeting, EU domain, actual EU usage.
3. TIER DETERMINATION (per service)
[ ] Intermediary — mere conduit / caching / hosting
[ ] Hosting — do we store information provided by a user?
(comments, reviews, uploads, bios, forums)
[ ] Online platform — store AND disseminate to the public?
[ ] Marketplace — consumers conclude distance contracts with
traders?
[ ] Small/micro enterprise exemption applies? Y / N
Headcount ____ Turnover ____ Balance sheet ____
** Recheck date: ____________ **
[ ] Approaching 45m average monthly active EU recipients?
4. OBLIGATIONS THAT FOLLOW
[list by tier, cumulative]
5. GAPS TODAY
[what we do not have]
6. RECOMMENDED SEQUENCE AND OWNER
Prepared by: ______ Reviewed by: ______ Circulated: ______
Annotation. Section 2's note exists because the first reaction inside every American company is that the rules cannot apply without European operations. Putting the correct rule in the template, in bold, ends that conversation on page one instead of in week six. Section 3's recheck date exists because the small enterprise exemption is real, valuable, and lost by growth that nobody re-examines.
Tool 2 — Legal representative selection memo
LEGAL REPRESENTATIVE — MEMBER STATE SELECTION
CANDIDATE MEMBER STATES
+-------------+------------------+------------------+----------+
| State | Coordinator | Activity level / | Language |
| | | resourcing | practice |
+-------------+------------------+------------------+----------+
FACTORS
[ ] Coordinator resourcing and enforcement posture
[ ] Language of proceedings and of required communications
[ ] Existing group presence or advisers in the state
[ ] Local counsel availability and cost
[ ] Practical experience of peers in our sector
CANDIDATE REPRESENTATIVE DILIGENCE
[ ] Can it actually receive and act on authority communications?
[ ] Does it have capacity, or is it a mail-forwarding service?
[ ] Does it understand it can be held liable for non-compliance?
[ ] Contract terms: scope, escalation, SLA, notice, termination
[ ] Insurance and indemnity
RECOMMENDATION: ______________ Approved: ______ Date: ______
FOLLOW-ON
[ ] Representative details published in the terms and on the site
[ ] Coordinator of establishment identified: ______________
[ ] Languages of communication published: ______________
Annotation. This decision is routinely made by whoever answers a vendor email first, and it determines which regulator supervises the company for years. The diligence block matters because the provider remains responsible for non-compliance regardless of what the representative does or fails to do — a mail drop transfers no risk.
Tool 3 — Terms and conditions clause set
CONTENT MODERATION — WHAT WE DO
We may restrict content or accounts as described below. This
section describes the policies, procedures, measures, and tools
we use, including algorithmic decision-making and human review.
Restrictions we may apply: removal · disabling access · reducing
visibility · restricting monetization · suspending or terminating
the service · suspending or terminating the account.
Grounds: (a) the content is illegal under applicable law; or
(b) the content is incompatible with [Policy], reproduced at
[link].
How decisions are made: [describe intake, triage, automated
detection tools and what they do, human review stages, who
decides, escalation]. We apply these measures diligently,
objectively, and proportionately, with due regard to the rights
and legitimate interests of all parties involved, including
fundamental rights.
Automated tools: we use [describe] to [detect / prioritize /
action]. Where a decision is made or supported by automated
means, we say so in the statement of reasons we send you.
NOTIFYING US ABOUT CONTENT
Anyone may notify us of content they believe is illegal, using
the form at [link]. [Summarize what a notice needs.] We
acknowledge receipt and tell you our decision and how to
challenge it.
IF WE RESTRICT YOUR CONTENT OR ACCOUNT
We will send you a statement of reasons explaining what we did,
why, on what legal or contractual ground, whether automated means
were used, the territorial scope and duration, and how to
challenge the decision.
CHALLENGING A DECISION
Internal complaints: free, available for six months, at [link].
Decisions are reviewed by qualified staff and not solely by
automated means. If your complaint has sufficient grounds we will
reverse the decision.
Out-of-court dispute settlement: you may select any certified
dispute settlement body. We will engage with it in good faith.
Judicial redress remains available.
MISUSE
After a prior warning, we may suspend, for a reasonable period,
users who frequently provide manifestly illegal content, and the
processing of notices or complaints from those who frequently
submit manifestly unfounded ones. Examples: [ ].
RECOMMENDATIONS
Where we rank or recommend content, the main parameters are
[list, most significant first, with why each matters]. You can
change [ ] at [link].
CONTACT
Users: [electronic contact, not solely automated]
Authorities: [electronic point of contact]
Languages: [ ]
Legal representative in the EU: [name, address, contact]
Annotation. The heading of the first clause — "what we do" rather than "rights we reserve" — is the whole point. Standard platform terms describe discretion; the regulation requires describing process. Drafting this clause forces the company to write down what the trust and safety team actually does, which is where the practices that should change get found.
Tool 4 — Notice submission form
REPORT ILLEGAL CONTENT
1. Where is the content?
Exact URL(s): ________________________________ [REQUIRED]
[Multiple URLs permitted — one per line]
2. What is the content?
[ ] Post [ ] Listing [ ] Review [ ] Message [ ] Media
[ ] Account [ ] Advertisement [ ] Other: ______
3. Why do you believe it is illegal?
Category: [dropdown — IP infringement · defamation · consumer
protection · counterfeit goods · product safety · data
protection · hate speech · CSAM · terrorist content · other]
Explanation (be specific — what law, and why this content
breaks it):
______________________________________________________
______________________________________________________ [REQUIRED]
4. Supporting material
[upload — e.g. proof of rights, court order, correspondence]
5. About you
Name: ______________ Email: ______________ [REQUIRED except
for notices concerning certain offences — see note]
Acting on behalf of: ______________
[ ] I am a trusted flagger. Designating authority: ______
6. Declaration [REQUIRED]
[ ] I confirm my good-faith belief that the information and
allegations in this notice are accurate and complete.
We will acknowledge receipt without undue delay, tell you our
decision and the reasons for it, and explain how to challenge it.
We will tell you whether automated means were used in processing
and deciding on your notice.
Annotation. Every field marked REQUIRED corresponds to an element that makes a notice legally sufficient. That is deliberate and it cuts both ways: a notice that is precise and adequately substantiated enough for a diligent provider to identify illegality without a detailed legal examination confers actual knowledge and starts the clock on the conditional liability exemption. A form that collects less produces notices that may not confer knowledge — but also produces notices you cannot act on. Collect the full set. Note also field 5's trusted flagger checkbox: it is what makes priority routing possible.
Tool 5 — Statement of reasons template and generation spec
STATEMENT OF REASONS
To: [recipient] Ref: [case ID] Date: [ ]
WHAT WE DID
Measure: [removed / disabled access to / reduced visibility of /
restricted monetization of / suspended / terminated] the
following: [content ref or account].
Territorial scope: [worldwide / EU / member state(s): ____]
Duration: [permanent / until ____ / ____ days]
WHY
Ground: [ ] The content is illegal.
[ ] The content is incompatible with our terms.
IF ILLEGAL:
Legal ground: [instrument and provision]
Why this content is illegal on that ground: [specific to THIS
content — not a category label]
IF TERMS:
Contractual ground: [clause reference and quoted text]
Why this content is incompatible: [specific to THIS content]
HOW WE DECIDED
This decision followed:
[ ] a notice submitted to us on [date] [by [notifier] — include
identity only where strictly necessary]
[ ] our own-initiative investigation
[ ] an order from [authority] dated [date]
Automated means were used:
[ ] in detecting or identifying the content
[ ] in making or supporting the decision
[ ] not used
[Describe briefly what the automated means did.]
HOW TO CHALLENGE THIS
1. Internal complaint — free, for six months from this date:
[link]. Reviewed by qualified staff, not solely automated.
2. Out-of-court dispute settlement — you may select any
certified body. We will engage in good faith: [info link]
3. Judicial redress in the courts of your member state.
[Company] · [contact] · [legal representative details]
Engineering generation specification:
TRIGGER: any enforcement action of type
REMOVE | DISABLE | DEMOTE | DEMONETIZE | SUSPEND | TERMINATE
where basis IN (ILLEGAL, TERMS)
REQUIRED FIELDS FROM THE CASE RECORD (enforce at write time —
a case cannot be actioned without them):
measure_type · territorial_scope · duration
basis (ILLEGAL | TERMS)
legal_ground_ref (required if ILLEGAL)
legal_ground_explanation (free text, required, min length)
terms_clause_ref (required if TERMS)
terms_explanation (free text, required, min length)
source (NOTICE | OWN_INITIATIVE | AUTHORITY_ORDER)
notifier_identity_disclosed (bool + justification if true)
automated_detection (bool) · automated_decision (bool)
automated_description
decided_by · decided_at
ACTIONS ON TRIGGER (atomic with the enforcement action):
1. Render statement of reasons
2. Deliver to affected recipient
3. Submit to the Commission transparency database
(strip personal data)
4. Open the complaint window (6 months)
5. Write to the audit log
EXCEPTION: skip where content is deceptive high-volume commercial
content.
Annotation. The generation specification is the single most valuable item in this toolkit, and the critical design decision is in the third line of REQUIRED FIELDS: a case cannot be actioned without them. If the explanation fields are optional, moderators will leave them blank under volume pressure and the company will publish thousands of non-compliant statements into a public database. Make the fields blocking. It slows moderation slightly and it is the difference between a compliant program and a visible failure.
Tool 6 — Complaint form and decision record
CHALLENGE A DECISION Case ref: [prefilled]
What was decided: [prefilled from the statement of reasons]
Date: [prefilled]
Why do you think this was wrong?
[ ] The content is not illegal
[ ] The content does not breach the policy cited
[ ] The decision relied on incorrect facts
[ ] The measure is disproportionate
[ ] Other: ______
Explain: ______________________________________________
Supporting material: [upload]
Contact: ______________
COMPLAINT DECISION RECORD (internal)
Case ref: ______ Original decision by: ______ Date: ______
Complaint received: ______ ** 6-month window opened: ______ **
Reviewer: ______ ** Different from original decision-maker? Y/N **
Reviewer qualification: ______________
Automated means used in this review: [ ] none [ ] supporting only
** Decision must NOT be solely automated **
Original ground reviewed: ______________
Does the complaint contain sufficient grounds to reverse? Y / N
Reasoning: ______________________________________________
DECISION: [ ] Upheld [ ] Reversed [ ] Modified: ______
Communicated to complainant: ______ (date)
Out-of-court settlement and judicial redress explained: [ ]
Median-time metric updated: [ ]
Reversal logged against: policy ______ reviewer ______
Annotation. The two boxed lines in the decision record are the compliance-critical ones: a different reviewer, and a decision not taken solely by automated means. The final line — logging reversals against both the policy and the reviewer — is the management value. A policy with a high reversal rate is a badly written policy; a reviewer far off the team median needs training. That data exists nowhere else.
Tool 7 — Trader verification file
TRADER VERIFICATION Trader ID: ______
INFORMATION OBTAINED (all required before trading is permitted)
[ ] Name
[ ] Address
[ ] Telephone
[ ] Email
[ ] Identification document / electronic identification — type
and reference: ______________
[ ] Payment account details
[ ] Trade register and registration number (where registered),
or equivalent means of identification: ______________
[ ] SELF-CERTIFICATION: commitment to offer only products and
services complying with applicable Union law [signed ____]
BEST-EFFORTS ASSESSMENT OF RELIABILITY
Sources used:
[ ] Official online database: ______________ Result: ______
[ ] Official interface: ______________ Result: ______
[ ] Supporting documents requested: ______________
Assessment outcome: [ ] Reliable [ ] Inconsistent — details ____
Assessed by: ______ Date: ______
IF INFORMATION IS INACCURATE / INCOMPLETE / OUT OF DATE
[ ] Correction requested (date: ______)
[ ] ** TRADER SUSPENDED pending correction ** (date: ______)
[ ] Corrected and reinstated (date: ______)
STORAGE
[ ] Held securely; retention = relationship + 6 months
[ ] Retention date: ______
RE-VERIFICATION (existing seller base)
Tier: [ ] High volume [ ] Mid [ ] Low [ ] Dormant
Notified: ______ Deadline: ______ Completed: ______
Outcome: [ ] Verified [ ] Suspended [ ] Departed
Annotation. The suspension line is the one companies resist and the one the regulation requires: where information appears inaccurate or incomplete and is not corrected, the trader must be suspended. Build the suspension as an automated consequence of a stale verification record rather than a manual decision someone has to make about a revenue-generating seller, because manual decisions of that kind do not get made.
Tool 8 — Advertising and recommender disclosures
AD DISCLOSURE — REQUIRED ELEMENTS, VISIBLE IN REAL TIME
[ ] Prominent marking that this is an advertisement
[ ] "Paid for by [payer]"
[ ] "Presented on behalf of [beneficiary]" (if different)
[ ] "Why am I seeing this?" → main targeting parameters, in
meaningful terms, and how to change them where applicable
PROHIBITIONS — VERIFY IN THE AD STACK
[ ] No profiling using special categories of personal data
[ ] No profiling-based ads where we are aware with reasonable
certainty the recipient is a minor
Note: this turns on what we KNOW, not on age verification.
Document what signals we have and how they are used.
RECOMMENDER EXPLANATION — IN THE TERMS AND IN-PRODUCT
The main parameters we use to rank and recommend, most
significant first:
1. ______________ Why it matters: ______________
2. ______________ Why it matters: ______________
3. ______________ Why it matters: ______________
What you can change: ______________
Where: [link, directly accessible from the ranked section]
Annotation. The minor-advertising prohibition is the one most often misread as an age-verification mandate. It is not: it prohibits profiling-based advertising where the provider is aware with reasonable certainty that the recipient is a minor. The compliance artifact is therefore a document describing what age signals the company holds and how the ad stack uses them — not an age-gate.
Tool 9 — Dark patterns audit
INTERFACE AUDIT — DECEPTIVE OR MANIPULATIVE DESIGN
Screen / flow reviewed: ______________ Reviewer: ______
[ ] Pre-checked boxes for consents, subscriptions, or add-ons
[ ] Asymmetric choice — "Accept" prominent, "Reject" buried or
requiring more clicks
[ ] Confirmshaming — decline options worded to induce guilt
[ ] False urgency — countdown timers that reset; "only N left"
not tied to real inventory
[ ] Obstruction — cancellation or deletion harder than signup
[ ] Nagging — repeated prompts after a decision
[ ] Default settings that favor us over the user's evident
preference
[ ] Hidden costs revealed late in a flow
[ ] Visual interference — low-contrast or de-emphasized options
[ ] Bait and switch
FINDINGS
Issue | Screen | Severity | Owner | Fix by | Fixed
------|--------|----------|-------|--------|------
Sign-off: ______ Date: ______ Next audit: ______
Annotation. Run this on the whole funnel, not just the consent screen, and run it with someone who did not design the flows. Every first audit finds something; a countdown timer that resets on page reload and a pre-checked box are the two most common. Note that several of these patterns are independently actionable under consumer protection law in both the Union and the United States — FTC Act § 5 reaches deceptive practices — so the audit has value well beyond this regime.
Tool 10 — Authority order response
[Date]
[Issuing authority]
Re: Order dated [date], reference [____]
Dear [ ],
[Company] acknowledges receipt on [date] of the above order
[to act against illegal content / to provide information].
EFFECT GIVEN
[Company] gave effect to the order on [date/time] as follows:
[describe precisely — what was removed/disabled/provided, scope].
[OR, where the order is deficient:]
Before giving effect, [Company] respectfully notes that the order
does not appear to contain [the statement of reasons / the exact
electronic location / information on redress / a statement of
territorial scope] required of such orders. [Company] requests
that the order be supplemented so that it can give effect
appropriately.
[OR, where scope is at issue:]
[Company] notes that the territorial scope of an order is limited
to what is strictly necessary to achieve its objective. The order
appears to seek [global] removal. [Company] has given effect
within [the Union / member state], and would welcome
clarification if a wider scope is intended.
RECIPIENT NOTIFIED
[For orders to act:] [Company] informed the recipient concerned
of the order and the effect given on [date].
CONTACT
[Company]'s point of contact for authorities is [ ], reachable
at [ ]. Communications may be addressed in [languages].
Respectfully,
Annotation. The three alternative middle paragraphs are the point. Companies default to silent compliance with any order that arrives, including deficient ones and ones claiming global scope. The regulation prescribes what an order must contain and limits territorial scope to what is strictly necessary — and a polite, specific request for supplementation is a normal supervisory interaction, not defiance.
Tool 11 — Transparency report template
[COMPANY] — DSA TRANSPARENCY REPORT — [Period]
1. ORDERS FROM MEMBER STATE AUTHORITIES
By type (act against illegal content / provide information),
by member state, by category of illegal content.
Median time to acknowledge: ____ Median time to give effect: ____
2. NOTICES VIA THE NOTICE AND ACTION MECHANISM
Total: ____
By category of alleged illegal content: [table]
Submitted by trusted flaggers: ____
Action taken: removed ____ · disabled ____ · demoted ____ ·
demonetized ____ · account action ____ · no action ____
Basis: law ____ / terms ____
Automated means used in processing: ____
Median processing time: ____
3. OWN-INITIATIVE CONTENT MODERATION
Measures taken, by type and by category.
Use of automated tools: [describe]
Numbers of items and accounts affected.
4. INTERNAL COMPLAINTS
Received: ____ By basis: [table]
Decisions: upheld ____ / reversed ____ / modified ____
** Reversal rate: ____% **
Median handling time: ____
5. AUTOMATED MEANS — QUALITATIVE DESCRIPTION
Purposes · indicators of accuracy · possible rate of error ·
safeguards applied.
6. OUT-OF-COURT DISPUTE SETTLEMENT
Disputes submitted: ____ Outcomes: ____ Median time: ____
Whether we implemented the decisions: ____
7. MISUSE SUSPENSIONS
Users suspended: ____ Notifiers/complainants suspended: ____
8. METHODOLOGY
Counting conventions. What counts as an "action". How median
times are calculated. Period covered.
Published in machine-readable format at: [link]
Annotation. Section 8 is not optional in practice even where it is not prescribed in terms: two companies counting "actions" differently produce numbers that cannot be compared, and a report without a methodology note invites the assumption that the numbers were chosen. The boxed reversal rate is the number readers will look at first, and it is better to publish it with an explanation than to have it computed by someone else.
Tool 12 — DMA business user rights letter
[Date]
[Gatekeeper] — Compliance Function
Re: Business user rights under the Digital Markets Act —
[Company], [account/developer/seller ID]
Dear [ ],
[Company] is a business user of [core platform service], a core
platform service provided by [Gatekeeper] as a designated
gatekeeper. We write to request the following, to which business
users are entitled.
1. DATA ACCESS. Effective, high-quality, continuous, and
real-time access to the aggregated and non-aggregated data
provided for or generated in the context of our use of the
service, including data provided or generated by end users
engaging with our products or services. Please confirm the
available access mechanism and provide credentials.
2. STEERING. Confirmation that we may communicate and promote
offers, including at different prices or conditions, to end
users acquired via the service, and conclude contracts with
those end users, whether or not they use the service, free of
charge and without restriction.
3. PRICING PARITY. Confirmation that we are not required to
offer the same prices or conditions on the service as we offer
through other channels.
4. ANCILLARY SERVICES. Confirmation that access is not
conditioned on our use of [Gatekeeper]'s identification
service, browser engine, or payment service.
5. ADVERTISING TRANSPARENCY. For advertising services: daily
information, free of charge, on each advertisement placed —
the price and fees paid by us, the remuneration received by the
publisher, and the metrics on which each is calculated. Please
confirm the access mechanism.
6. NON-USE OF OUR DATA. Confirmation that non-public data
generated by our activity on the service is not used by
[Gatekeeper] in competition with us.
7. ACCESS CONDITIONS. Confirmation that the general conditions
of access applicable to us are fair, reasonable, and
non-discriminatory, and a copy of those conditions.
Please respond by [date]. We would prefer to resolve these matters
directly and reserve all rights, including recourse to the
European Commission and national authorities.
Very truly yours,
Annotation. This is the only tool in the toolkit that can make a company money, and almost nobody sends it. The seven items are entitlements, not requests, and the compliance functions on the receiving end exist precisely to process them. Assign this file to a commercial owner — the person who runs the platform relationship — rather than to compliance, because the value is in using the data and the freedoms, not in filing the response.
Tool 13 — Geo-blocking assessment
SHOULD WE SERVE THE EU AT ALL?
THE NUMBERS
EU users: ____ (____% of total) EU revenue: $____ (____%)
3-year EU forecast: ______________
One-time build cost: $__________
Annual operating cost: $__________
(moderation and complaint staffing · statement-of-reasons
pipeline · representative · dispute settlement · reporting)
EU revenue per compliance dollar: ______
IF WE STAY
[ ] Full program per the implementation plan
[ ] Harmonize globally / geo-differentiate — decided: ______
IF WE EXIT — DO IT PROPERLY
[ ] Actual technical geo-blocking, not a banner
[ ] Terms prohibit use from the Union
[ ] No EU-language interface or marketing
[ ] No euro pricing
[ ] No EU-specific payment methods
[ ] No EU country domains
[ ] No EU-targeted advertising
[ ] Existing EU users notified, with data export and refunds
NOTE: the test is whether the service is OFFERED to recipients
in the Union. Geo-blocking while marketing in French and pricing
in euros creates a record of having tried, not an exit.
REVISIT: ____________ (annually — year-two costs fall materially)
Annotation. Include this tool in the program even when the answer is obviously to comply. A board that has seen the alternative priced approves the compliance budget differently from one that has only seen a cost. And for a company with genuinely marginal European exposure, exiting is a rational business decision rather than a failure — provided it is done as a real exit rather than a gesture.
Related documents
- The Digital Services Act and Digital Markets Act: What US Companies Actually Have to Do
- Complying With the DSA and DMA: A Practical Guide
- DSA and DMA Readiness Checklist: A Practical Checklist
- Platform Liability Toolkit: Terms of Service, Notice Procedures, and Litigation Defenses
- Running a Platform That Hosts User Content: A Practical Guide
- Antitrust for Technology Companies: Monopolization, Platform Conduct, and the IP-Antitrust Interface
This toolkit is general information, not legal advice, and does not create an attorney-client relationship.
