The deliverable is that nothing happened

Proofreading services for documents that come back the same
and better only where something was actually wrong.

Every other service here makes something. A page argues, copy persuades, a manual instructs. A proofread document comes back substantively identical, and its success is measured by restraint rather than by what was added.

Errors corrected and nothing else. Claims flagged rather than altered. Names left exactly as found, however wrong they look. Your style guide given precedence over our judgment, and your own pattern given precedence over both. Every change recorded, so you can reject any of them.

WHAT THE WORK IS 01 / 04

A proofread document reads exactly like one nobody touched. That is the problem, and it is why every change is recorded.

  • 01 The faultClassified first.
  • 02 The ownerDecides the action.
  • 03 The listNothing silent.
Claims flagged, never altered Names left as found Your guide over our judgment Every change reversible

The Writely difference

A proofreader reads the document again and cannot help improving it. That is the entire hazard of this job.

Improvement is not proofreading. Every other service on this site is about making the text better, and this one is about not making it different, which is why a change here has to be defensible as the correction of an identified fault rather than as a preference.

Classify the fault before deciding what to do about it

A fault is not a fault is not a fault, and the class decides the action. A doubled word is plainly ours to remove. A comma may not be an error at all, and may be a decision the house style has already settled. A figure that disagrees with the table three pages earlier is a real find, but the fix is to raise it, because correcting it silently makes the document assert something its author never checked. And a claim about what a study found belongs to the author in a way a grammar error never does. Classifying first is what stops a competent reader from making a large number of small confident interventions the author did not ask for.

Flag every factual problem and change none of them

This is the firmest boundary on the page and the one most often got wrong by people who proofread. If the text says four hundred participants and the figure elsewhere says forty, the correction is obvious and it still must not be made silently, because the proofread document would then state something nobody verified. It goes on the flag list with its location, and the author decides. The same applies to names, dates, places, totals and anything a reference points to: we check whether the document agrees with itself, not whether the world agrees with the document. A client who wants the facts checked is asking for fact-checking, a different service with a different method.

Never normalise a name that looks wrong

The most damaging quiet change in this document type is the one where a name gets corrected into something more familiar. Institutional names are not always in a dictionary. Town names vary between spellings. Surnames in this part of the world are frequently compound and frequently unfamiliar, and traditional titles are translated rather than transcribed. All of these are correct as they stand, and a reader's instinct that they are wrong is precisely the reason they get changed. So proper nouns are never touched. They are checked for consistency instead, because a document that calls the same body two different things on pages four and nine has a real fault, and that is a different problem from a name that looks unfamiliar.

Your guide and your own pattern outrank our judgment

There is no single correct English, which is why style guides exist and why anyone who tells you a comma is wrong without saying which guide they are working from is not telling you anything useful. So the guide the client supplies decides, including its silences, and where it says nothing the author's own established pattern decides, because it is the one the rest of the document follows. That is the difference between a document that reads like the rest of the author's work and one that reads as though a stranger had been through it, and it is why the style note records where the guide ran out.

Find what your document has not settled

Eight things a proofreading job has to decide, and the one with the best prose has decided four of them by accident.

None of this is about being strict. It is about the document coming back as the same document, with the faults gone and nothing else altered.

Agree in writing what is in scope before anything starts

Proofreading, editing and rewriting are three different jobs with three different scopes and three different outcomes, and most disappointing proofreading is a disagreement about which one was ordered rather than a disagreement about quality. So the scope is written down first: which of the three, which style guide, whether references are included, whether anything may be restructured, and what happens to the flag list. That takes one paragraph and it removes the single most common source of an awkward conversation at the end. Where the honest answer turns out to be the larger job, content audits is the page that covers what proofreading will not do on purpose.

Correct the error, not the sentence around it

The distinction that keeps proofreading apart from editing. If a word is wrong, the word is replaced and the sentence stays as it was, including its rhythm, its length and its awkwardness if it had any. Rewriting the sentence to accommodate the correction turns a one-word intervention into a change of voice, and a document full of those is one the author has to read again from the start. It also removes the possibility of checking the work, because there is no longer a list of corrections to review. A change is either the correction of a specific fault or it is not made.

Check the document against itself

The pass that finds more than any other and requires no knowledge of the subject at all. The same term should not appear as two words. The same figure should not appear as two values. A source cited in the text should be in the list and the other way round. A table should agree with the paragraph above it. None of that needs an argument and none of it is a matter of taste, which makes it the most defensible part of the job, and it is where an unqualified proofreader adds the most value per hour.

Be consistent where both choices are defensible

Two defensible readings in the same document is a fault even though neither is wrong, because the inconsistency is what a reader notices and what makes a document read as careless. So once a choice is made anywhere in the document it holds throughout, whether it is a spelling, a capitalisation, a form of address, a numbering style, a date format or a way of presenting a measurement. The decisions come from the author's own earliest usage rather than from preference, and the choice is recorded in the style note so that the next person to touch the document can see it was on purpose. A document that is unremarkable but internally consistent reads as more reliable than a slightly better written one that contradicts itself on page twenty.

Fix the document you have, not the one you intended

A persistent source of argument in this work is a reference to a section or table that the author is certain exists and which is not in the document that was actually submitted, usually because a revision moved it and one cross-reference was missed. The proofreader cannot know which of the two is right, so the reference is flagged rather than re-pointed at the nearest plausible target. Guessing here produces a document that looks finished and is wrong, and it is wrong in a way nobody will find until somebody follows the reference. The same applies to a figure mentioned in the abstract that has vanished from the body, and it is a flag that takes the author a minute and saves them a retraction.

Leave the numbers, dates and units to the author

Internal consistency is checked and applied; anything else about a figure belongs to the author. So a total that does not add up is a flag rather than a correction, because adding it up ourselves means choosing a version of the document, and the arithmetic error is often in the source data rather than in the sum. The same goes for a currency that changes partway through a table, a date that is impossible, a unit applied to the wrong quantity, and a percentage that does not match the fraction above it. These are all flagged with their locations. It sounds cautious and it is the only defensible position, because a document that quietly agrees with itself is a document nobody can rely on.

Report what you did not change, not only what you did

The deliverable is a document plus two lists, and the second list is the one clients forget they are paying for. The change list has every correction with its original, so the author can accept or reject each one. The flag list has everything noticed and left alone on purpose: the factual questions, the names, the cross-references, the figures, and the places where the house style was silent. That list is the actual value of the job, because it is the part that needs a human decision and the part a reader could not have produced alone. It is also what makes the work reviewable by somebody other than the author, which matters when the author is not the person who will answer the questions.

What this service cannot do

Four limits. Nothing here is a claim about what is true: sources are not retrieved, their existence is not confirmed, and no claim in the document is checked against anything outside it. Nothing is rewritten, restructured or improved for style, because improvement is a different job. No proper noun is normalised, however wrong it looks, because unfamiliar is not the same as incorrect. And no change is made silently, because the failure this service exists to prevent is the one where a reader cannot tell what happened to their document.

If every change is defensible as a correction and everything else is flagged, the document has not been edited. That sentence is the whole job, and it is worth reading into every agreement before the work starts.

A better brief

Send the style guide, the scope, and the deadline the document actually has to meet.

The brief that stalls a proofreading job is the one that sends the file and asks for it back polished, because polished is not a scope and three different jobs answer to it. What settles the work is short: which of proofreading, editing and rewriting, which style guide and what it covers, whether references and tables are included, what the deadline really is, and who answers the flag list. Given that, the job is mechanical in the good sense, and the only judgement calls are the ones the brief has already made unnecessary.

  • Proofreading, editing or rewriting.
  • The style guide, and what it covers.
  • Whether references and tables are included.
  • Whether facts may be flagged, never changed.
  • The real deadline, and who signs off.
  • Who answers the flag list.
Share your brief
SCOPE CHECK Before starting

Six things before the file is opened.

Scope written down
Style guide supplied
Facts flagged only
Names untouched
Pattern decides
Two lists returned

How the work is done

From a finished draft to the same draft with the faults gone and two lists attached.

Agreement, first pass, consistency, return. The first step is the only one that requires a conversation, and it is the one that prevents every awkward one afterwards.

01 / THE AGREEMENT

Write down the scope, the guide, and who answers the flags

First, because every later judgement depends on it and because it is the step most often skipped. Which of the three jobs it is, which style guide applies and what it covers, whether references, tables and figures are in scope, and what the client wants done about factual problems, which this service will only flag. It also establishes who responds to the flag list and by when, since a flag with nobody to answer it is not useful and the deadline for the whole job usually depends on it. This step is short and it is what turns proofreading from a negotiation into a procedure, which is what the client is actually paying for.

02 / THE FIRST PASS

Read for faults only, and resist every improvement

Then the pass that matters most and requires the most discipline, because the reader is good and the document is nearly right, which is the worst conditions for restraint. Every fault is recorded with its location and its correction, classified as it is found so that the class decides whether it is fixed or raised. Anything noticed that is not a fault goes to the flag list rather than into the document. The test applied to each proposed change is whether the author would agree it was wrong, and if the honest answer is that both versions are defensible, it stays. Where you can see a sentence that would read better and you leave it alone, that is the job being done properly.

03 / THE CONSISTENCY PASS

Put the document against itself, term by term and figure by figure

Then the pass that needs no knowledge of the subject and finds the most. The same term under two names, the same figure under two values, a cross-reference to a section that is not there, a source cited in the text and missing from the list, an abbreviation expanded twice, a table that disagrees with the paragraph above it. Each is recorded with its locations and the pattern is fixed where the author is clearly right, and flagged where the author is not. That is why this pass is named apart from the first: it uses a different tool, it makes no stylistic changes at all, and it is the part of the job that is impossible to do well by reading for prose.

04 / THE RETURN

Send the file back with both lists, and mean it when you say nothing was changed

Finally the return, which is a document and two lists and nothing else. The change list has every correction with its original and its fault class, so the author can accept or reject each one individually. The flag list has everything noticed and left alone on purpose. The style note records which guide was used, where it said nothing, and which of the author's own patterns was followed instead. When nothing was changed beyond a handful of faults, that is said plainly rather than dressed up, because a proofreader who cannot report a clean document has not been doing the job. Where improvement is the point instead, persuasive copy editing is the opposite contract.

What actually gets handed over

Three documents, and the flag list is the one clients did not know they were paying for.

The proofread file is the obvious part. The other two are where the judgement is, and one of them is the reason this job is worth having rather than skipped.

The change list

Every correction made, in document order, with the original, the replacement and the class of fault. Nothing goes into the file that is not on this list, which is what makes the work reviewable: an author can accept or reject each entry individually rather than reading a document wondering what happened to it. It also makes the proofreader's own reasoning auditable, which is the only way a client can tell the difference between a correction and a preference without having to read as closely as the proofreader did. Where a change was made against a marginal preference rather than a clear fault, it says so on the list, so it can be rejected easily.

The flag list

Everything noticed and left alone on purpose, with locations: the figures that disagree with each other, the cross-reference pointing at nothing, the name spelled two ways, the abbreviation expanded twice, the citation that may or may not be in the list, the date that looks impossible. This is the document that carries the actual value, because each item needs a decision from somebody who knows the answer and none of them is a proofreading question. It is also the reason a reader could not have produced it alone, and the reason an author who has been working on a document for months tends to find the longest item on it is the one they had stopped seeing.

The style note

Short, and it settles most future arguments: which style guide was applied, what it covers, where it was silent, and which of the author's own established patterns was followed instead. Where the guide was silent and the document contained a genuine choice, the choice that was followed is recorded, so the next person to touch the file can see it was on purpose rather than an oversight. That matters more than it sounds, because the alternative is a decision being silently reversed by a later reader who assumed it was a mistake, which is exactly the kind of quiet change this service exists to prevent.

The scope record

The agreement, kept and attached: which of the three jobs this was, what was in and out, who answered the flag list and when, and what the deadline actually was. It is the only thing that resolves a disagreement later, because it settles what was asked rather than what was delivered, and the two are frequently confused by people looking at the file alone. In practice it is one paragraph and it prevents most of the conversations that proofreading jobs otherwise end in, because it records in advance that facts are flagged rather than checked and that names are not normalised.

The clean report

Included because it is unusual and because it is honest: a short statement of what was found and what was not, including the categories that came back empty. A file with two corrections in it is a file that was nearly right, and saying so plainly is more useful than delivering a list long enough to look thorough. It also gives the author a genuine picture of the document, which matters when the document is being submitted to somebody else and the author needs to know what was actually checked. Where nothing was checked because nothing needed it, that is the most useful sentence the whole job produces.

The flag list is the part clients keep. A corrected file goes out and is never seen again. The flags are the items that needed a human decision, and they are the reason a proofread document is worth having even when the file itself was nearly clean.

What a proofreading job has to settle

Six things, and the version with the best prose has settled four of them by accident rather than on purpose.

Each can be settled before the file is opened, and each decides whether the document comes back as the same document.

  • Scope written down: which of the three jobs.
  • Style guide applied, and its silences noted.
  • Facts flagged, never altered.
  • Proper nouns untouched, however unfamiliar.
  • Consistency established against itself.
  • Change list and flag list both returned.
THE JOB Six things

Six things a job must settle.

Scope in writing
Guide applied
Facts flagged only
Names untouched
Checked against itself
Both lists returned

Explore related academic support

The pages that produce the document you are about to have checked.

Proofreading comes last in a process that has to have happened first, and a file that was never drafted cannot be proofread, only read. These are the pages that produce the material this job checks, and they are worth using if the underlying problem is that the document does not exist yet.

Not sure which of the three jobs you need? Send the file and say what is wrong with it. The honest answer is frequently the smallest one, and saying so before the work starts is worth more than an hour of deciding afterwards.

Where the boundaries of this job are written down

A professional association leads, because its account of where proofreading stops and editing begins is the closest thing this trade has to a definition of its own edges.

Eighteen candidates were tried across two rounds and eight answered, and the failures were almost entirely the professional bodies for this kind of work, which says something about how the field is organised. What survived leads with that association, then two overviews that put proofreading and the larger job next to each other so the boundary has something to be measured against. Then two entries on style: one on why a guide exists given there is no single correct English, and one indexing the guides. Then the reference standard for decisions with an authority behind them, a dictionary for decisions about what a word means, and a writing laboratory for the distinction between the three jobs. All eight are new to this corpus. These external resources are informative references, not Writely service pages.

WHERE ONE JOB ENDS AND THE NEXT BEGINS

Copyediting.com

The professional association for this work, and the source for the distinction the whole page turns on. Their material on the difference between proofreading and editing, on what belongs to each, and on why the boundary is a matter of scope rather than of skill is the useful part, because it settles the argument that causes most disappointing proofreading jobs. Also useful on working with writers, on the agreement that should precede the work, and on the standards a document is held to before it goes anywhere.

Open resource
WHAT THE JOB ACTUALLY IS

Wikipedia: Proofreading

The orientation piece, and useful precisely because it records that the word covers several quite different practices depending on the field and the era. Its sections on the history of the term, on the difference between proofing a printed page and marking a text, and on the techniques involved are all directly relevant to setting expectations before a job starts. Worth reading alongside the copyediting entry below, since the two together are the clearest short account of where the work divides.

Open resource
THE JOB IMMEDIATELY ABOVE THIS ONE

Wikipedia: Copyediting

Included because the boundary on this page is only meaningful against it. Copyediting reaches sentence construction, clarity, tone and sometimes structure, which is a different intervention with a different risk profile and a different review method. Its sections on the scope of the work, on the disputes within the profession about how far to go, and on the criticisms that have been aimed at both jobs are useful for a client deciding which one they are buying.

Open resource
WHY A GUIDE IS NEEDED AT ALL

Wikipedia: Style Guide

The page that makes the argument this service relies on: there is no single correct English, which is why style guides exist and why a person who tells you a comma is wrong without naming a guide has told you nothing. Its material on the purposes a guide serves, on how one is written, and on the tensions between an institution and its individual writers is the background to the rule that your guide outranks our judgment. Also useful on what a guide is for when it is silent.

Open resource
THE PRACTICAL ANSWER TO ANY DISPUTE

Wikipedia: List of Style Guides

An index rather than a guide, and the most immediately useful thing on this page for settling a question in thirty seconds. It shows how many authorities there are, that they disagree with each other on points most people argue about, and that publishers, universities, journals and professions each maintain their own. That last point is the argument for the page's position in a single line: a document is proofread against the guide its own institution follows, not against whichever one the reader happens to know.

Open resource
THE DECISIONS THAT HAVE AN AUTHORITY BEHIND THEM

Chicago Manual of Style Online

Included for the narrow set of questions where a document is either right or wrong against something rather than being a matter of taste: how references are presented, how quotations are marked, what an attribution carries, how numbers and dates are set. Those are the checks in a proofreading job that nobody should be arguing about, because an authority exists. Also useful on the parts of style that are genuinely left to the author, which is the part of it worth reading before deciding what to be strict about.

Open resource
WHAT A WORD ACTUALLY MEANS

Cambridge Dictionary

Included because a great many editing disputes are settled by what a word means rather than by what sounds better, and by both regional forms where a word is spelled differently across the Atlantic. Useful for the check that stops a confident correction of something that was correct, and for the distinction between a word used wrongly and a word used in an unusual register. It is also where the question of whether a particular usage is standard or merely common gets an answer rather than an opinion.

Open resource
THE THREE JOBS, DISTINGUISHED

Purdue OWL

The writing laboratory, useful for the plainest account of the difference between proofreading, copyediting and revision, which is the distinction a client needs before ordering rather than after. Their material on common errors, on the punctuation and usage questions that come up most, and on how to read comments on your own writing is the practical half of what this page is arguing for. Also useful as a source of the checks a document should get through before it is proofread at all.

Open resource

Names that are correct and look wrong

Nigeria at the centre. The house guide wins.

Writely is rooted in Nigeria and works remotely with students, researchers, institutions and businesses. Whether the document is a thesis, an article, a report, a contract or a set of policy notes, and whether you are in Lagos, Abuja, a campus somewhere, or another part of the world entirely, you can work on it here.

Remote proofreading and editing support

Proofreading across Nigeria

Explore proofreading for students, researchers, institutions and businesses in Nigerian cities, communities and states. Three things are genuinely different here and all three belong in the brief. The first is names. Institutional names in this part of the world are not always in a dictionary, town names vary between spellings and between official and customary forms, surnames are frequently compound, and traditional titles are translated rather than written out. All of these are correct as they stand, and every one of them is something an editor working outside the country will want to fix, which is why the rule on this page is that proper nouns are never touched. The second is that the house style is chosen rather than universal: an institution has one, a journal has another, and the same author publishes under both, so the guide supplied outranks any general authority. The third is that contested usages in institutional and administrative writing here are settled by the institution rather than by a dictionary, which is a matter of asking rather than of guessing.

Lagos Abuja Port Harcourt Kano Ibadan Enugu Benin City Uyo Kaduna Calabar Jos Abeokuta Awka Akure Warri Owerri Zaria Sokoto Bauchi Maiduguri

Nigeria states and the Federal Capital Territory

Writely’s support is available remotely across Nigeria, including:

  • Abia
  • Adamawa
  • Akwa Ibom
  • Anambra
  • Bauchi
  • Bayelsa
  • Benue
  • Borno
  • Cross River
  • Delta
  • Ebonyi
  • Edo
  • Ekiti
  • Enugu
  • Gombe
  • Imo
  • Jigawa
  • Kaduna
  • Kano
  • Katsina
  • Kebbi
  • Kogi
  • Kwara
  • Lagos
  • Nasarawa
  • Niger
  • Ogun
  • Ondo
  • Osun
  • Oyo
  • Plateau
  • Rivers
  • Sokoto
  • Taraba
  • Yobe
  • Zamfara
  • Federal Capital Territory

Remote support across Africa

Country-level discovery

Writely welcomes conversations with students, researchers and institutions across Africa, including in:

Nigeria Algeria Angola Benin Botswana Burkina Faso Burundi Cabo Verde Cameroon Central African Republic Chad Comoros Côte d’Ivory Democratic Republic of the Congo Djibouti Egypt Equatorial Guinea Eritrea Eswatini Ethiopia Gabon Gambia Ghana Guinea Guinea-Bissau Kenya Lesotho Liberia Libya Madagascar Malawi Mali Mauritania Mauritius Morocco Mozambique Namibia Niger Rwanda São Tomé and Príncipe Senegal Seychelles Sierra Leone Somalia South Africa South Sudan Sudan Tanzania Togo Tunisia Uganda Western Sahara Zambia Zimbabwe

Questions, answered

Start with the scope.

If your question is not covered here, send the essentials in a chat. Which of the three jobs you want settles more than anything else you could tell us.

Ask your question
What is proofreading, and how is it different from copyediting?
Proofreading finds and corrects faults. Copyediting additionally reaches sentence construction, clarity, tone and sometimes structure, and rewriting goes further still and can change what the document says. The difference matters because they are three different jobs with three different scopes, and most disappointing proofreading is a disagreement about which one was ordered rather than about quality. In proofreading the sentence stays as it was and the error in it is fixed; in copyediting the sentence may be rebuilt. That is why the scope gets written down before the file is opened, because it is the one thing that cannot be settled afterwards.
What will you change and what will you leave alone?
Changed: typos, missing or duplicated words, spelling and punctuation that your guide has settled, agreement and tense that leave the meaning intact, and inconsistencies where the document contradicts itself, such as the same term under two names or a cross-reference pointing at nothing. Left alone: anything about whether a claim is true, every proper noun however unfamiliar it looks, any sentence that is merely less good than it could be, and every question of taste where both options are defensible. The test applied to each change is whether you would agree it was wrong, and if the honest answer is that you might not, it goes on the flag list instead.
Do you check whether the facts are right?
No, and this is the firmest limit on the page. We check whether the document agrees with itself, not whether the world agrees with the document. A figure that disagrees with another figure is a genuine find and it goes on the flag list with its location; it is not corrected, because a silently corrected figure makes the document say something nobody verified. Sources are not retrieved and their existence is not confirmed. If you want the facts checked against outside material, that is fact-checking: a different service, a different method, and a person who can be held to the answer.
Will you use our style guide?
Yes, and it decides. The guide you supply outranks any general authority, and where it is silent your own established pattern in the document decides, which is often better evidence than a guide because the rest of the file follows it. Both decisions are recorded in the style note so the next person to touch the document can see they were on purpose. Where your convention and our preference differ, yours wins without argument, because a document that reads like the rest of your work is worth more than one that reads as though a visitor had been through it.
Can you tell me exactly what you changed?
Yes, and that is part of the deliverable rather than an extra. You get the proofread file, a change list with every correction, its original and the class of fault, and a flag list of everything noticed and left alone on purpose with its location. You can accept or reject each change individually instead of reading the file wondering what happened to it. Where a change was made against a marginal preference rather than a clear fault, it says so on the list so it is easy to reject.
What do you need from me before you start?
The file, the style guide if you have one, the scope in a sentence, and the deadline that is real rather than the one you hope for. If there is a submission guideline or a template from a journal or institution, that is the most useful thing you can send, because it settles most of the questions before they arise. It also helps to say who will answer the flag list, since several of those items need somebody who knows the answer and the schedule of the whole job often depends on it.
How long does proofreading take?
Long enough to do the first pass and the consistency pass properly, which are two distinct readings rather than one fast one, and the consistency pass is where most of the finds come from. Speed comes from scope rather than from hurrying: a document with references and tables takes longer than one without, and a file that has been rewritten several times usually needs the most work because it has accumulated layers. If the deadline is genuinely immovable, say so before the work starts rather than after, because the honest options are a narrower scope or a later date, and both are better than a file that was read too quickly.
Can you work in a document with tracked changes?
Yes, and it is usually the best format for this work, because every correction arrives as a visible change that can be accepted or rejected individually, which is exactly the review model the job requires. The change list still comes on its own, since tracked changes show what was done and not why, and the distinction between a correction and a preference is the part that matters to you. Whatever format is used, you keep the original throughout and nothing is altered in a way you cannot undo.
What languages do you work in?
English, and the house conventions that go with it in this part of the world, including the spellings and forms of address that Nigerian institutional and legal writing uses and that an editor working from another market will otherwise get wrong. Beyond English, documents in other languages can be checked for internal consistency, since a term that appears as two words is a fault in any language, but they are not proofread to a standard here. For that, the right person is a native speaker of the language who also understands what the document is for, and we will say so rather than attempt it.
Do you proofread theses, articles and reports?
All three, and they want slightly different things. A thesis wants its structure, its referencing and its figures checked more carefully than its sentences. An article wants its abstract, its in-text citations and its reference list to agree with each other and with the journal's style, which is where most of the finds are. A report wants its tables to agree with its prose, its figures to be consistent, and its names to be spelled the same way throughout. The three are more alike than clients expect, which is why the scope question at the start settles most of it.
Do you check references and citation format?
Yes, if they are in scope, and they are the part of a document most worth checking. We confirm that every source cited in the text appears in the list, that every entry in the list is cited, that the details agree between the two places, and that the format follows the guide you supplied or the one your institution uses. We do not retrieve the sources and we do not confirm that a work exists, which is a different kind of check altogether. Where an entry looks incomplete we flag it rather than completing it, because guessing an author or a year is worse than leaving the gap visible.
Where is Writely available, and how do I start?
Writely is based in Nigeria and works remotely with students, researchers, institutions and businesses across Nigeria and the rest of Africa. Use the Chat with us button to send the file, the style guide and one sentence on the scope. That is enough to tell you honestly which of the three jobs you need, which is frequently the smallest answer.

Your next move

Send the file, the guide and the scope. We’ll send back the same document and two lists.

Whether you need the faults out, the consistency established, the facts flagged rather than touched, or the names left exactly as found, begin a clear conversation with Writely.

Request services