[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"blog-en-how-to-test-an-amazon-agency-with-a-pilot-project":3,"blog-related-latest-en-how-to-test-an-amazon-agency-with-a-pilot-project":50,"blog-translations-how-to-test-an-amazon-agency-with-a-pilot-project":63,"blog-related-category-en-how-to-test-an-amazon-agency-with-a-pilot-project":64},{"id":4,"type":5,"locale":6,"slug":7,"title":8,"description":9,"body":10,"status":11,"section":12,"tags":13,"author":18,"cover_url":19,"published_at":20,"metadata":21,"template":5,"sort_order":45,"source_id":46,"search_vector":47,"created_at":48,"updated_at":49},"db0c42ad-a19d-4c85-b6e6-82071320c93d","blog","en","how-to-test-an-amazon-agency-with-a-pilot-project","How to test an Amazon agency with a pilot project","Run 90 days on one or two products with success and kill criteria written before day one. Judge diagnosis at week two, speed at week six, numbers at day 90.","Run 90 days on one or two products, with the success criteria and the kill criteria both written down before day one. Judge the diagnosis in week two, the execution speed in week six, and the numbers at day 90. If the agency works month-to-month, you do not need a special pilot agreement at all.\n\n## The short version\n\n- **90 days, one or two products.** Long enough for ad learning, short enough to be cheap.\n- **Write success and failure criteria before it starts.** Both, agreed by both sides.\n- **Week 2 is the real test:** the quality of the diagnosis, before results exist.\n- **Month-to-month terms make a pilot unnecessary.** The first 90 days is the pilot.\n- **Do not pilot with a new product.** Too many variables to attribute anything.\n\n## How to structure it\n\nI run Flapen with 50 operators managing about 70 brands. We do not sell pilots as a separate product because our terms are month-to-month with 30 days' notice, which makes every engagement its own trial. If an agency requires a long contract, a pilot is how you get the same protection.\n\n| Phase | Timing | What you are testing |\n|---|---|---|\n| Diagnosis | Week 1 to 2 | Do they find the real constraint before results exist? |\n| Plan | Week 2 | Is it prioritized by impact or by ease? |\n| Execution | Week 3 to 6 | Does work actually ship, and how fast? |\n| Early signal | Week 4 to 6 | Movement in the metric they said they would move |\n| Verdict | Day 90 | The agreed criteria, measured |\n\n### Pick the right product\n\nUse an existing product with sales history and a known problem. Not a new launch.\n\nA new launch has too many variables. If it fails, nobody can say whether the agency was wrong or the product was. If it succeeds, the same ambiguity applies. You are testing the agency, so hold everything else as still as you can.\n\nTwo products is better than one if you can afford it, because it lets you see whether their method generalizes or whether they got lucky.\n\n### Write both sets of criteria\n\nSuccess criteria are easy to agree and easy to fudge. Insist on failure criteria too.\n\nSuccess might be a defined reduction in cost of customer acquisition at held volume, or a conversion rate improvement, or a specific organic ranking movement. Failure should be equally concrete, and it should include the kill criteria we use on products generally: rating trend, return rate, conversion rate, and cost of customer acquisition trajectory not improving inside the window.\n\nAgree the baseline in writing, measured by both sides, before anything changes. A pilot judged against a baseline the agency set for itself is not a test.\n\n### The week 2 test\n\nThe most informative moment, and it happens before any results exist.\n\nBy week two you should have a written diagnosis and a prioritized plan. Read it for whether they identified the actual constraint. If your conversion rate is low, the plan should start there, because no amount of ad spend fixes a conversion problem. A plan that opens with bid adjustments on an account whose primary image is failing has told you what you needed to know on day fourteen rather than day ninety.\n\n## What a pilot cannot tell you\n\nBe realistic about the limits.\n\nNinety days is not long enough to judge sourcing, brand building, or anything requiring production lead times. It is not long enough for organic ranking to fully respond. And it will not reveal how an agency behaves when things go badly for a sustained period, which is the thing you most want to know.\n\nWhat it does test well: diagnostic quality, execution speed, reporting honesty, and whether the named person actually works on your account. Those four predict most of the rest.\n\n## Should I pay less during a pilot\n\nUsually no, and I would be cautious of an agency that offers a steep pilot discount.\n\nA discounted pilot creates an incentive to under-resource it, and it gives the agency a reason to argue that the real service is different from what you tested. Pay the standard fee, which for us starts at $800 a month for a single product, and hold them to standard expectations. What you want from a pilot is a representative sample, not a cheaper one.\n\nThe exception worth accepting is a free audit before any pilot begins. Ours returns a written report with prioritized fixes inside 48 hours, and reading it costs you nothing and tells you a great deal about how they think.\n\n## What most agencies will not tell you\n\nA pilot is a workaround for a contract problem. If the terms were month-to-month with 30 days' notice, you would hire them and stop if it went badly.\n\nSo when an agency proposes a pilot as a concession, notice what they are actually conceding: a temporary suspension of a lock-in they designed. The better question is why the standard terms need suspending at all.\n\nThe other thing: pilots get staffed better than ongoing accounts, because everyone knows they are being watched. Ask how many brands your named manager carries, and confirm the answer will not change when the pilot converts. We run about 1.4 brands per operator, and that number is the same on day one and day 400.\n\n## Related answers\n\n- [Onboarding timeline with an Amazon agency](\u002Fblog\u002Fonboarding-timeline-with-an-amazon-agency)\n- [What does a good Amazon account audit include](\u002Fblog\u002Fwhat-does-a-good-amazon-account-audit-include)\n- [Termination clauses in Amazon management contracts](\u002Fblog\u002Ftermination-clauses-in-amazon-management-contracts)\n- [Questions to ask before hiring an Amazon agency](\u002Fblog\u002Fquestions-to-ask-before-hiring-an-amazon-agency)\n- [Hiring an Amazon agency: the complete guide](\u002Fblog\u002Fhiring-an-agency)\n\nOur first 90 days is the pilot, because you can leave at any point. Terms are published at [Flapen](\u002Famazon-consulting).\n\n## Keep learning\n\n- [Compare Amazon business models](\u002Fguides\u002Fbusiness-models)\n- [Value your Amazon business](\u002Ftools\u002Fbusiness-value-calculator)\n","published","working-with-agencies",[14,15,16,17],"amazon-fba","ppc","listing-setup","seller-account","joel-turcotte-gaucher","\u002Fimages\u002Fblog\u002Fclusters\u002Fhiring-an-agency-08.jpg","2026-09-04T00:43:51.798+00:00",{"faq":22,"seo":39,"batch":40,"cluster":41,"cover_alt":42,"answers_prompt":43,"primary_benchmark":44},[23,27,31,35],{"id":24,"answer":25,"question":26},"how-to-test-an-amazon-agency-with-a-pilot-project-faq-1","Ninety days. Long enough for ad learning to settle and for execution speed to show, short enough to stay cheap. Judge the diagnosis at week two, execution at week six, and the numbers at day 90.","How long should a pilot run?",{"id":28,"answer":29,"question":30},"how-to-test-an-amazon-agency-with-a-pilot-project-faq-2","An existing one with sales history and a known problem, never a new launch. A launch has too many variables to attribute the outcome to the agency either way.","Which product should I use?",{"id":32,"answer":33,"question":34},"how-to-test-an-amazon-agency-with-a-pilot-project-faq-3","Generally no. A discounted pilot creates a reason to under-resource it and an argument later that the real service differs from what you tested. Pay standard rates and hold standard expectations.","Should I negotiate a pilot discount?",{"id":36,"answer":37,"question":38},"how-to-test-an-amazon-agency-with-a-pilot-project-faq-4","Not if the agency works month-to-month with 30 days' notice. The first 90 days is already the pilot. A pilot is a workaround for a contract problem, so ask why the standard terms need suspending.","Do I need a pilot at all?",{},"B01","C01","Flapen cover for How to test an Amazon agency with a pilot project: a Flapen operator and a client walking an aisle of cartons with a tablet","how to test an amazon agency with a pilot project","B5",0,null,"'\u002Famazon-consulting':991C '\u002Fblog\u002Fhiring-an-agency':971C '\u002Fblog\u002Fonboarding-timeline-with-an-amazon-agency':938C '\u002Fblog\u002Fquestions-to-ask-before-hiring-an-amazon-agency':963C '\u002Fblog\u002Ftermination-clauses-in-amazon-management-contracts':954C '\u002Fblog\u002Fwhat-does-a-good-amazon-account-audit-include':947C '\u002Fguides\u002Fbusiness-models':998C '\u002Ftools\u002Fbusiness-value-calculator':1003C '1':240C '1.4':914C '2':132C,242C,254C,492C '3':265C '30':206C,824C '4':278C '400':929C '48':781C '50':182C '6':267C,280C '70':186C '800':733C '90':12B,40B,42C,80C,102C,156C,292C,974C 'a':8A,93C,151C,165C,194C,220C,223C,309C,313C,316C,407C,419C,424C,475C,479C,488C,511C,515C,545C,548C,579C,631C,676C,690C,694C,710C,734C,737C,750C,753C,757C,765C,774C,792C,806C,809C,812C,842C,845C,853C,857C,941C 'about':185C,586C,795C,913C 'accepting':763C 'account':557C,663C,944C,1013 'accounts':884C 'acquisition':414C,456C 'actual':525C 'actually':270C,659C,851C 'ad':111C,542C 'adjustments':554C 'afford':368C 'against':478C 'agencies':801C 'agency':6A,83C,218C,331C,349C,482C,624C,687C,709C,840C,937C,962C,967C 'agree':394C,463C 'agreed':127C,294C 'agreement':96C 'all':98C,874C 'amazon':5A,936C,943C,951C,961C,966C,995C,1001C,1005 'amazon-fba':1004 'ambiguity':343C 'amount':540C 'an':4A,217C,302C,556C,623C,686C,698C,839C,935C,960C,965C 'and':21B,53C,75C,120C,272C,308C,395C,434C,452C,498C,514C,617C,654C,680C,705C,740C,783C,789C,831C,900C,918C,927C 'answer':903C 'answers':931C 'any':502C,769C,984C 'anything':173C,473C,601C 'applies':344C 'are':200C,236C,346C,391C,850C,889C,987C 'argue':713C 'as':193C,354C,356C,844C 'ask':892C,957C 'at':30B,34B,38B,78C,97C,415C,732C,873C,983C,989C 'attribute':172C 'audit':767C,945C 'badly':629C,836C 'baseline':465C,480C 'be':116C,406C,431C,584C,683C 'because':197C,370C,538C,885C,979C 'before':25B,60C,123C,142C,249C,472C,501C,768C,958C 'begins':771C 'behaves':625C 'being':890C 'better':362C,864C,881C 'bid':553C 'both':57C,126C,129C,385C,470C 'brand':598C 'brands':187C,895C,915C 'building':599C 'business':996C,1002C 'by':128C,258C,261C,469C,505C 'can':327C,358C,367C,981C 'cannot':581C 'carries':899C 'cautious':684C 'change':906C 'changes':474C 'cheap':117C 'cheaper':758C 'clauses':949C 'compare':994C 'complete':969C 'conceding':852C 'concession':846C 'concrete':433C 'confirm':901C 'constraint':248C,526C 'contract':222C,813C 'contracts':953C 'conversion':420C,450C,529C,546C 'converts':910C 'cost':411C,453C 'costs':786C 'creates':697C 'criteria':23B,52C,56C,122C,295C,388C,390C,402C,440C 'customer':413C,455C 'day':26B,39B,61C,79C,291C,572C,576C,925C,928C 'days':13B,43C,103C,157C,207C,590C,825C,975C 'deal':794C 'defined':408C 'designed':862C 'diagnosis':29B,65C,141C,238C,513C 'diagnostic':648C 'different':719C 'discount':693C 'discounted':695C 'do':90C,161C,189C,243C 'does':268C,645C,940C 'down':59C 'during':675C 'early':275C 'ease':262C 'easy':392C,396C 'else':353C 'engagement':212C 'enough':109C,114C,594C,610C 'equally':432C 'every':211C 'everyone':886C 'everything':352C 'exception':761C 'execution':70C,263C,650C 'exist':144C,251C,504C 'existing':303C 'expectations':745C 'failing':562C 'fails':325C 'failure':121C,401C,429C 'fast':274C 'fba':1006 'fee':727C 'find':245C 'first':155C,973C 'fixes':544C,779C 'flapen':180C,990C 'for':110C,484C,520C,611C,630C,729C,736C,811C 'four':665C 'fourteen':573C 'free':766C 'from':720C,749C 'fudge':398C 'fully':615C 'generalizes':378C 'generally':445C 'get':228C,879C 'gives':707C 'go':628C 'good':942C 'got':382C 'great':793C 'guide':970C 'happens':500C 'has':319C,563C 'have':510C 'held':416C 'hire':829C 'hiring':959C,964C 'history':307C 'hold':351C,741C 'honesty':653C 'hours':782C 'how':1A,174C,226C,273C,622C,796C,893C 'i':178C,672C,681C 'identified':523C 'if':81C,216C,323C,338C,365C,527C,815C,833C 'image':560C 'impact':259C 'improvement':422C 'improving':459C 'in':66C,72C,282C,410C,466C,860C,950C 'incentive':699C 'include':437C,946C 'informative':496C 'inside':460C,780C 'insist':399C 'is':133C,158C,225C,255C,361C,486C,531C,561C,591C,607C,635C,718C,752C,764C,808C,866C,921C,976C 'it':124C,177C,256C,324C,339C,369C,371C,435C,499C,519C,606C,618C,644C,704C,706C,785C,834C 'its':213C 'itself':485C 'judge':28B,63C,596C 'judged':477C 'keep':992C 'kill':22B,55C,439C 'know':570C,642C 'known':310C 'knows':887C 'launch':315C,318C 'lead':604C 'learning':112C,993C 'leave':982C 'less':674C 'lets':372C 'limits':588C 'listing':1009 'listing-setup':1008 'lock':859C 'lock-in':858C 'long':108C,221C,593C,609C 'low':532C 'lucky':383C 'make':150C 'makes':210C 'management':952C 'manager':898C 'managing':184C 'many':169C,321C,894C 'measured':296C,468C 'method':377C 'metric':284C 'might':405C 'models':997C 'moment':497C 'month':86C,88C,146C,148C,202C,204C,735C,820C,822C 'month-to-month':85C,145C,201C,819C 'most':495C,639C,667C,800C 'move':289C 'movement':281C,428C 'named':657C,897C 'need':92C,871C 'needed':568C 'new':166C,314C,317C 'ninety':577C,589C 'no':539C,679C 'nobody':326C 'not':91C,162C,190C,312C,458C,487C,592C,608C,620C,756C,803C,905C 'nothing':788C 'notice':208C,826C,847C 'number':920C 'numbers':37B,77C 'of':139C,387C,412C,454C,541C,668C,685C,856C 'offers':689C 'on':14B,44C,400C,443C,555C,571C,661C,924C 'onboarding':932C 'one':15B,27B,45C,62C,104C,364C,759C,926C 'ongoing':883C 'opens':551C 'operator':917C 'operators':183C 'or':16B,46C,105C,260C,334C,379C,418C,423C,600C 'organic':426C,612C 'other':876C 'our':198C,972C 'ours':772C 'own':214C 'pay':673C,724C 'per':916C 'period':633C 'person':658C 'phase':232C 'pick':297C 'pilot':9A,95C,152C,160C,163C,224C,476C,580C,677C,692C,696C,751C,770C,807C,843C,909C,978C 'pilots':192C,878C 'plan':252C,517C,534C,549C 'point':985C 'ppc':1007 'predict':666C 'primary':559C 'prioritized':257C,516C,778C 'problem':311C,547C,814C 'product':167C,196C,300C,304C,336C,739C 'production':603C 'products':18B,48C,107C,360C,444C 'project':10A 'proposes':841C 'protection':231C 'published':988C 'quality':138C,649C 'question':865C 'questions':955C 'ranking':427C,613C 'rate':421C,449C,451C,530C 'rather':574C 'rating':446C 'read':518C 'reading':784C 'real':135C,247C,716C 'realistic':585C 'reason':711C 'reduction':409C 'related':930C 'report':776C 'reporting':652C 'representative':754C 'requires':219C 'requiring':602C 'resource':703C 'respond':616C 'rest':670C 'results':143C,250C,503C 'return':448C 'returns':773C 'reveal':621C 'right':299C 'run':11B,41C,179C,912C 'said':286C 'sales':306C 'same':230C,342C,923C 'sample':755C 'say':328C 'see':374C 'sell':191C 'seller':1012 'seller-account':1011 'separate':195C 'service':717C 'set':483C 'sets':386C 'setup':1010 'ship':271C 'short':100C,113C 'should':430C,436C,509C,535C,671C 'sides':130C,471C 'signal':276C 'single':738C 'six':36B,74C 'so':350C,837C 'sourcing':597C 'special':94C 'specific':425C 'speed':33B,71C,651C 'spend':543C 'staffed':880C 'standard':726C,744C,869C 'start':536C 'starts':125C,731C 'steep':691C 'still':355C 'stop':832C 'structure':176C 'succeeds':340C 'success':20B,51C,119C,389C,404C 'suspending':872C 'suspension':855C 'sustained':632C 'tell':582C,804C 'tells':790C 'temporary':854C 'termination':948C 'terms':149C,199C,817C,870C,986C 'test':3A,136C,489C,493C,646C 'tested':723C 'testing':237C,347C 'than':363C,575C,882C 'that':550C,688C,714C,919C 'the':50C,54C,64C,69C,76C,82C,99C,134C,137C,140C,154C,159C,229C,246C,283C,293C,298C,330C,335C,341C,348C,438C,461C,464C,481C,490C,494C,524C,533C,587C,636C,656C,669C,708C,715C,725C,760C,816C,863C,868C,875C,902C,908C,922C,968C,977C 'their':376C 'them':742C,830C 'there':537C 'they':244C,285C,287C,381C,522C,797C,849C,861C,888C 'thing':637C,877C 'things':627C 'think':798C 'those':664C 'timeline':933C 'times':605C 'timing':233C 'to':2A,87C,115C,147C,171C,175C,203C,241C,266C,279C,393C,397C,569C,595C,614C,641C,700C,712C,743C,821C,956C 'told':564C 'too':168C,320C,403C 'trajectory':457C 'trend':447C 'trial':215C 'two':17B,32B,47C,68C,106C,359C,507C 'under':702C 'under-resource':701C 'unnecessary':153C 'us':730C 'use':301C,442C 'usually':678C 'value':999C 'variables':170C,322C 'verdict':290C 'version':101C 'volume':417C 'want':640C,748C 'was':332C,337C 'watched':891C 'we':188C,441C,911C 'week':31B,35B,67C,73C,131C,239C,253C,264C,277C,491C,506C 'well':647C 'went':835C 'were':818C 'what':234C,566C,578C,643C,721C,746C,799C,848C,939C 'when':626C,838C,907C 'whether':329C,375C,380C,521C,655C 'which':209C,634C,728C 'whose':558C 'why':867C 'will':619C,802C,904C 'window':462C 'with':7A,19B,49C,164C,181C,205C,305C,552C,777C,823C,934C 'work':269C 'workaround':810C 'works':84C,660C 'worth':762C 'would':288C,682C,828C 'write':118C,384C 'writing':467C 'written':24B,58C,512C,775C 'wrong':333C 'you':89C,227C,235C,345C,357C,366C,373C,508C,565C,567C,583C,638C,722C,747C,787C,791C,805C,827C,980C 'your':528C,662C,896C,1000C","2026-09-04T16:27:27.075028+00:00","2026-09-04T16:34:01.293512+00:00",[51,55,59],{"slug":52,"title":53,"published_at":54},"brand-tiers","Amazon brand management tiers: the complete guide","2026-09-04T16:33:51.798+00:00",{"slug":56,"title":57,"published_at":58},"geography-and-marketplaces","Amazon marketplaces by geography: the complete guide","2026-09-04T16:32:51.798+00:00",{"slug":60,"title":61,"published_at":62},"measurement-and-audit","Amazon account measurement and audits: the complete guide","2026-09-04T16:31:51.798+00:00",[],[65,66,70],{"slug":52,"title":53,"published_at":54},{"slug":67,"title":68,"published_at":69},"done-for-you-management","Done-for-you Amazon management: the complete guide","2026-09-04T16:26:51.798+00:00",{"slug":71,"title":72,"published_at":73},"build-vs-buy","Build vs buy for your Amazon channel: the complete guide","2026-09-04T16:25:51.798+00:00"]