If you have typed "LinkedIn automation tool" into Google this quarter, you already know the problem. Half the results pitch you a product, the other half warn you about bans, and nobody actually explains how the channel works in 2026. This guide is the reference I wish existed when we started Flow AI: what LinkedIn automation is, what is safe, what moves reply rates, and how to build a system around it that books meetings without burning accounts.
What is LinkedIn automation?
LinkedIn automation is any software that performs actions on LinkedIn on your behalf. In the broadest sense that includes scraping profiles, auto-viewing pages, sending connection requests, sending direct messages, liking posts, following up on unanswered threads, and piping replies into a CRM. In practice, most teams use the term to mean one specific thing: a tool that runs an outbound outreach sequence from one or more LinkedIn accounts, so a salesperson, founder, recruiter, or agency does not have to click through every step manually.
There are three rough categories of LinkedIn automation on the market today. The first is the browser extension, which hijacks your logged-in session and performs actions from your machine. It is cheap, often buggy, and the category most likely to trip LinkedIn's detection. The second is the cloud tool, which logs into your account from a server and runs actions server-side. It is more scalable, and it is also the category LinkedIn has spent the most effort trying to catch out. The third, which is where Flow AI sits, is a managed outreach platform: a cloud tool that does the automation, plus everything around it (search, warm-up, multi-sender rotation, a shared inbox, CRM, and AI drafting) so the automation is the boring part rather than the main event.
What LinkedIn automation is not is "send a pitch to 10,000 strangers on autopilot and wait for a flood of meetings." That era is over. LinkedIn's detection systems, the platform's per-account limits, and buyer saturation have all moved in the same direction: if your message could have been sent to anyone, it will get ignored by everyone. In 2026, automation earns its keep when it compresses the admin around a good motion, not when it blasts volume.
So when I talk about LinkedIn automation in this guide, I mean the following stack of jobs:
- Finding the right prospects with filters that do not drown you in noise.
- Warming up profiles so a connection request lands on a page the prospect has seen before.
- Sending connection requests at a cadence one human could plausibly hit.
- Drafting a first message that reads like a person wrote it.
- Following up at the right tempo without pestering.
- Keeping the replies in one place so handoffs and metrics are honest.
If a tool helps with all six, it is doing LinkedIn automation properly. If it only does step three at volume, you have bought yourself a problem. The rest of this guide takes each job in turn.
Is LinkedIn automation safe in 2026?
Short answer: yes, if you treat LinkedIn like a platform you want to keep using. No, if you treat it like a commodity inbox. LinkedIn does not publish its full detection logic, but after running more than 10,000 accounts through Flow AI, the pattern is consistent. Accounts that get restricted are almost always the ones that break one of three rules: too much volume for the account's age, too repetitive a pattern of actions, or too many abandoned artefacts like outstanding connection requests. Accounts that stay healthy do the opposite.
The safe version of automation in 2026 looks like this. You cap connection requests at roughly 15 per day per account, which is where LinkedIn's informal ceiling sits and where Flow AI's Autopilot lands by default. You space actions out across a 9am to 6pm window in the account's local time zone, in short bursts roughly 15 minutes apart, so the log looks like a working day rather than a cron job. You visit a profile and engage with a recent post before you send the request, because humans do this by accident and bots usually do not. You run brand-new accounts through a 15-day warm-up so you start at one request on day one and climb to the full fifteen by day fifteen. And you withdraw connection requests that are still outstanding after 21 days, because an account with dozens of ignored invites looks suspicious fast.
I also treat fully hands-off automation as a design mistake, even when it is technically safe. The moment you try to auto-send the first message, auto-follow-up, auto-close, the quality of the conversation collapses. I wrote about the specific failure modes in Fully automated LinkedIn outreach, and why it rarely works, but the short version is that the parts of LinkedIn outreach that win are the parts a human has to do: pick the right 20 prospects, write a compliment that could not be copy-pasted to anyone else, and answer a reply like a peer rather than a sales rep.
The safest way to think about it: automate the work that is repetitive and low-judgment (searching, warming up, visiting profiles, sending connections, logging replies). Keep a human on the work that is high-judgment (the first message, the reply, the call invite). We go much deeper on the mechanics in How to automate LinkedIn outreach without getting banned, and on the warm-up curve itself in LinkedIn account warm-up.
If you want the deep dive on what LinkedIn's detection systems actually look for, what a restriction feels like, and how to recover if you have already tripped one, the dedicated safety sub-guide covers every scenario we have seen across our customer base. Treat this section as the headline, and that page as the full service manual.
Personalization vs templates
The single biggest lever on your reply rate is not your subject line, your cadence, or your tool. It is whether the first line of your message could have been copy-pasted to anyone else on your list. If the answer is yes, you are one of the thousand "Hi [Name]" notes the prospect ignored this week. If the answer is no, you have bought yourself five seconds of attention, which is almost all you need.
Real personalization does not mean ten paragraphs of flattery. It means one specific line that proves you looked. A reference to a post they wrote last month. A detail from their About section. A mutual thing you noticed in their career history. That is it. In the Outreach Playbook, step three is literally called "Create the ice-breaker" and the formula is two or three sentences long: compliment, then qualifying question. Nothing more.
The classic mistake I see is "personalization at scale" where the personalization is a merge tag (their company name, their city) and the rest of the message is generic. Prospects parse that out instantly. You get the open-rate of a template without the warmth of a real message, which is worse than either. If you want to see what good personalization looks like on the page, Personalize LinkedIn outreach at scale walks through the worked examples we run with customers.
Where AI helps, and where it does not, is the interesting part. AI is extremely good at reading a prospect's profile and recent activity and suggesting a compliment you can edit in 10 seconds. It is not good at understanding what is interesting about a specific person. Flow AI's Co-pilot is built for exactly this split: it reads the prospect's profile, their recent posts, your conversation history, and your offer context, then it drafts a reply in your voice. You keep the last edit. Nothing sends without your approval. The result is that the grind of writing 30 openers a day drops to a few minutes, and the quality stays high because a human still signs off.
If you are worried about AI messages sounding like AI, that is a real and solvable problem. The fix is two rules. Rule one: never ship the first draft. Rule two: strip the opening line if it starts with "I hope this finds you well" or anything similar. We go into more detail on the voice problem in How to make AI LinkedIn replies sound like you. The takeaway for this section is simple: templates at scale are dead, personalization at scale is possible, and the way to get there is AI drafting plus human editing, not one or the other.
Multi-sender outreach
Once you have the message right, the next bottleneck is volume per account. LinkedIn caps each account at roughly 15 connection requests per day when you want to stay healthy. That is a hard ceiling. If your target is 100 conversations a week, one sender cannot get there no matter how good the software is. Multi-sender outreach is how teams solve the math.
The idea is straightforward. Instead of running one account at 15 requests per day, you run five accounts at 15 requests per day, split the prospect list between them, and suddenly you are at 75 requests per day and 525 a week without pushing any single account past its limit. This is the standard approach for agencies running outreach for clients, for sales teams with multiple AEs, and for founders who have put their executive team on the channel.
The thing almost nobody explains is that multi-sender is easy to describe and hard to run well. The pitfalls, in order: two senders prospecting the same contact and looking like a coordinated spam ring; replies landing in five different inboxes so nobody responds in time; reporting that cannot tell you which sender is driving which meeting; and handoff confusion when a deal goes to a closer. Flow AI's Multiple Senders feature handles the first and last of those by rotating prospects across assigned accounts with no overlap and making sender ownership visible everywhere. The unified inbox solves the second. And the dashboard solves the third by letting you filter every metric by sender or list.
I broke down the mechanics in detail in Multi-sender LinkedIn campaigns at scale and the team-flow view in Multi-sender outreach, the honest version. The one thing I will add here is a warning: do not run multi-sender if you have not run single-sender first. Multi-sender multiplies whatever you feed it. If the message is weak, a team of five will produce five times as many ignored messages. If the message is strong, they will produce five times as many conversations. Single-sender first, multi-sender second.
For agencies, multi-sender is really the core of the job. One agency, many client accounts, often dozens of sender profiles, all needing their own cadence and reporting. We built a dedicated LinkedIn automation for agencies sub-guide that walks through workspace structure, client onboarding, and billing for exactly this shape of team. If you are running outreach for more than two clients, that is the page to read next.
LinkedIn connection request limits in 2026
Let me give you the numbers we actually use, because the internet is full of stale ones. LinkedIn's informal ceiling for connection requests is around 15 per day for an established account. Brand-new accounts need to start lower. Sales Navigator accounts get slightly more room but not the 100 a day some old articles still claim. If you push above 15 consistently, the risk of a restriction rises in a non-linear way, and you do not get a warning.
There is also a weekly dimension that is less well understood. LinkedIn has been known to apply invitation throttles at the account level (roughly 100 to 200 a week) that kick in before the daily cap ever does. That is why spreading sends across the working week matters more than it used to. You do not want to do all 75 of your weekly sends on a Monday and then be silent for four days. The platform reads "Monday only" as not-a-human behaviour.
Three rules I give every customer the first time they set up. First, stay at or under 15 requests per day per account, even if the UI technically lets you send more. Second, spread sends across business hours in the local time zone of the sender, not the prospect. 9am to 6pm is the default window in Flow AI for exactly this reason. Third, withdraw connection requests that are still pending after 21 days. A pile of old, ignored invites is one of the cleanest signals to LinkedIn that the account is running a pattern rather than having actual conversations. Flow AI's Autopilot withdraws these automatically; if you are running your own tool, add a reminder in your calendar to prune them.
If you want the full picture (why the limits exist, what triggers a restriction, and how to recover), I wrote the long version in LinkedIn connection request limits in 2026. For this guide, the headline is: 15 per day per account, spread across the week, no orphans left over 21 days. Build your volume plan on that number and you will be in the clear. Try to squeeze more and you are playing with an account you cannot easily replace.
The other limit worth flagging is on profile activity. Flow AI's default ceiling is 80 post likes and 80 profile visits per account per day, which is well within what a power user would do naturally. You do not need to be anywhere near those numbers to run effective campaigns, but it is useful to know they exist so you understand the margin of safety in the product.
Account warm-up
Warm-up is the unglamorous part of automation that saves you from a restriction in month two. The idea is simple: a brand-new LinkedIn account that has never sent a connection request before cannot wake up on Monday and send 15 requests on Tuesday without looking suspicious. It needs a ramp.
The curve we use inside Flow AI is linear over 15 days. Day one, one request. Day two, two requests. Day three, three. And so on, up to day 15, which is the full 15 per day from then on. That slope is long enough to let LinkedIn's systems settle, short enough that the account is at full capacity inside three weeks, and predictable enough that a team running multiple accounts knows exactly when each one goes live.
There is more to a warm-up than just the send cadence, though. For a truly new account (an employee who has never been active), you want to round out the profile first: a real headshot, a headline that names the role and the company, an About section that reads like prose rather than a résumé, and ideally one or two recent posts or comments so the profile does not look dormant. Sales Navigator counts here too. An account that has had Sales Navigator for a while looks more established to LinkedIn's systems than one that just signed up. None of these are strictly required, but every one of them reduces the probability of a flag in the first two weeks.
For accounts you are moving from "active but manual" to "active with automation", the warm-up is slightly different. You are not introducing the account to LinkedIn; you are introducing a new pattern of activity. In that case I still start at roughly 5 requests per day and climb to 15 over a week, just to give the platform a chance to notice the change without reacting to it. Flow AI handles this automatically, but the principle applies whatever tool you use.
The deep version of this is in LinkedIn account warm-up, the 15-day curve. The summary for this guide: do not skip the ramp, do not compress the 15 days into 5, and do not try to run a brand-new account at full volume because "the other ones are warmed up". Every account has its own curve. Treat them that way.
Cold vs warm approach
Cold outreach means messaging people who have never heard of you. Warm outreach means messaging people who already have some touchpoint with you: a post they engaged with, a profile they viewed, a follow they started. Every real LinkedIn program uses both, but the ratio changes the results more than most teams expect.
In the Outreach Playbook I break prospects into three tiers by temperature. Hottest: people who viewed your profile in the last week. Warm: people who commented on or liked a recent post, commenters first. Lukewarm: new followers who have not engaged yet, plus whoever you pulled from Sales Navigator today. The rule is almost boring: work the hottest tier first, the warm tier second, the lukewarm tier third. If you skip that order and go straight to the Sales Navigator export, your acceptance rate drops and your reply rate drops further.
I do not love the word "warm" by itself because it hides what is actually going on. The thing that matters is whether the prospect has any reason to recognise you when the connection request lands. The best version of "warm" is that they saw a post of yours last Tuesday. The second best version is that the Autopilot sequence has quietly visited their profile and liked their most recent post in the days before the request, so when the notification appears, your name is not a cold start. That is exactly the engagement step baked into Flow AI's Autopilot sequence, and it is why the same message converts better through Flow than through tools that only send the connection request itself.
For a side-by-side on the mechanics and the message differences, Warm vs cold outreach on LinkedIn walks through real examples from Tom's account. The short version is that warm outreach trades volume for conversion: fewer messages, higher reply rate, better meetings. Cold outreach is the opposite, and pure cold on LinkedIn in 2026 is a bad bet. The middle path, which is what most teams actually run, is to use automation to manufacture a little bit of warmth (profile visit, post like) on top of every "cold" prospect before any message goes out.
If you are coming to this channel from email or cold calling and want a deeper comparison of how the motions differ, the automated vs manual LinkedIn outreach sub-guide covers the trade-off in more detail. For this section the takeaway is: temperature matters more than volume, and the easiest way to raise temperature is to engage with a prospect's content before you ever ask for a connection.
Sales Navigator filters
Sales Navigator is the single best piece of B2B targeting software LinkedIn has ever shipped. It is also the reason so many outreach programs fail. The filters are so powerful that teams build 200,000-person lists and then wonder why the replies are generic. The answer is: big lists make bad outreach. Sales Navigator is a scalpel, not a fishing net.
The filters I use most often, in rough order of impact: current job title, seniority, company headcount, industry, geography, and time in current role. Time in current role is the sleeper one. Someone who has been in a VP seat for nine months is dramatically more likely to engage with a new tool than someone who has been there for nine years, because they are still in the "building my stack" window. Combine that with a specific job title and a tight headcount band and you have a list of maybe 2,000 people that will reply at three to five times the rate of a 50,000-person broad list.
Flow AI's Search matches Sales Navigator's filters one for one, with two additions. First, there is a profile-keyword filter (like Apollo's) where you can require or exclude words in the headline, About section, or experience. That is the filter that catches people who describe themselves as "head of demand gen" in a sub-title even though their official title is Marketing Manager. Second, there is a natural-language search that converts a phrase like "VPs of product at Series B SaaS companies in the UK" into filters you can review and tweak. A backend filter also requires that a prospect has posted on LinkedIn in the last 30 days, which quietly lifts acceptance and reply rates by making sure the person is actually active. The full breakdown is on the Find Leads page.
On how to build a list well, the pattern I use with customers is in Sales Navigator filters for B2B. Short version: start with a narrow definition of who you want, not a broad one. Pull 500 to 2,000 prospects, not 20,000. Read the first 50 by eye before you add them to a campaign. If more than two or three of them look wrong for your offer, your filters are too broad and you need to tighten them before you spend a single credit.
One more thing. Do you actually need Sales Navigator? If you are running outreach seriously, yes. The basic LinkedIn search will not give you the filter depth to build a clean list. If you are experimenting, how to find your ideal buyers on LinkedIn without Sales Navigator has the manual workarounds. Most teams upgrade inside a month because the time saved pays for the seat on day one.
The first message
If you get one thing right in LinkedIn outreach, make it the first message. It is the bottleneck on everything downstream: acceptance rate, reply rate, call rate, close rate, all of it. The first message is also the single biggest thing automation cannot do for you. You have to write it. But you can make writing it much faster, and you can follow a formula that works.
The formula I use is in step three of the Outreach Playbook: two parts, compliment plus qualifying question. The compliment is specific (a detail from a post or profile), short (one or two lines), and casual. The qualifying question is a yes/no that names the outcome I help with. In full:
Hey [Name], [specific compliment they would care about]. Congrats :)
Outreach Playbook, step 3, first message template
How's [goal] progressing, is it going well?
Written out with a real example from the playbook: "Hey [Name], I enjoyed your recent post on scaling teams. I think we all wish we'd learned the 'hiring slow and firing fast' principle sooner :) How's hiring going for you and the team, are things progressing well?" Notice how short it is. Notice that there is no company name drop, no product pitch, no link, no ask for a call. The whole message fits on a phone screen without scrolling. That is the target.
Where the qualifying question changes is by offer. If I help with lead gen, I ask about lead gen. If I help with leadership coaching, I ask about team performance. The playbook shows a handful of variants, but the structure is always the same: "How's [the outcome I help with] progressing, is it going well?" The reason this works is that it is easy to answer, it surfaces whether there is a problem, and it naturally leads to a follow-up about a call if the answer is "not really". A longer breakdown with real screenshots is in The LinkedIn message that gets replies, and if the compliment half is the part you struggle with, LinkedIn ice-breakers without a pitch has 12 worked examples.
What the first message should not do: introduce your company, explain your product, drop a calendar link, or contain the word "synergy". If the prospect wants to know what you do, they will click your profile. That is why the headline on your profile matters so much. A first message that pitches is not shorter than your pitch; it is a waste of your one good chance at getting a reply. Hold the pitch for the call. Hold the link for the moment they ask for one. Everything you have to say fits in two sentences.
Follow-up sequences
Most replies are not to the first message. They are to the second or third touch, days or weeks later. This is the part of the channel that rewards patience, which is exactly what most teams run out of first. Set a rhythm you can stick to, and follow up twice before you move on.
The way I follow up has changed a lot over the years. "Just bringing this to the top of your inbox" used to work. It does not anymore. In 2026, that kind of nudge reads as pushy and lowers your credibility. The follow-ups that work now all have one thing in common: they provide something, even if it is small, before they ask for anything. A meaningful comment on the prospect's recent post. A voice note, which still stands out because so few people send them. A short video (camera on), which stands out even more. A link to a post or podcast you made that would actually help them. Or a simple follow-up that restates the outcome you help with, one week later, without the "nudge, nudge" energy.
Cadence-wise, I give it about a week between follow-ups. Too soon reads as desperate. Too late and the thread has gone cold. If they have not replied after two proper follow-ups, I move on and update their status in my pipeline. The inbox does not go away; sometimes people resurface a month later, and that is fine, because the follow-up I sent was the kind of thing they would be happy to see again rather than a guilt trip.
A specific Flow AI feature that matters here is scheduled LinkedIn DMs. Gmail-style scheduling for LinkedIn lets you write all three touches in one sitting (message, follow-up one, follow-up two), queue them for the right days, and move on. You do not need to remember a thread six days later. You also do not need to rewrite the follow-up in a panic on Friday afternoon. The full pattern, with examples of the voice notes and video follow-ups that have worked for us, is in LinkedIn follow-up messages that do not feel annoying.
The mindset shift is the important bit. Follow-ups are not reminders. They are additional openings. Each one should be a reason, on its own, for the prospect to want to reply. If it does not clear that bar, you are better off not sending it. Two well-made follow-ups beat five "circling back" notes every time.
Metrics that matter
LinkedIn outreach has a lot of numbers you can watch and only a few that tell you anything useful. I look at four, in this order: acceptance rate, reply rate, positive reply rate, and meetings booked. Everything else is noise or a derivative of one of those four.
Acceptance rate is connection requests accepted divided by connection requests sent. It tells you whether your targeting and your headline are right. A healthy acceptance rate for a well-targeted list is 30% to 40%. Below 20% and either your ICP is wrong or your profile looks like a pitch page. Above 50% and either your list is too warm (existing network) or your volume is too low to be meaningful.
Reply rate is replies divided by first messages sent. This tells you whether your first message is doing its job. A healthy reply rate with a specific, personalized opener is 15% to 25%. Templates will get you 3% to 5%. The difference is why personalization matters. Positive reply rate is replies that lead somewhere useful divided by total replies. "Not interested" is a reply; it is not a positive one. This is the number I watch most closely because it is the earliest signal that the message and the ICP are matched.
Meetings booked is meetings booked. There is no weighted version, no "intent-qualified" adjustment. It is the only number that matters to the pipeline. Everything upstream is interesting but optional. If meetings are flat, you do not have a campaign; you have a hobby.
Flow AI's dashboard shows all of these per list and per sender, filterable by time frame and segment, which matters more than it sounds. Aggregate numbers hide the story. The moment you slice by sender or by list, you see that one campaign is pulling the whole team up or one sender has a reply rate 3x the others, and that is where the coaching and the copying should go. The deeper breakdown is in LinkedIn outreach metrics that matter and the common misreadings (why your reply rate is "fine" but your pipeline is empty) are in LinkedIn outreach mistakes that wreck your reply rate.
Two more principles. Measure weekly, not daily. Daily numbers on small lists are noise; weekly numbers are signal. And run changes one at a time. Change the opener and the ICP in the same week, and you will never know which one moved the number. Change one, wait a week, then change the next. It feels slow, and it is the only way to actually learn the channel.
How to set up a LinkedIn automation system
Here is the six-step setup I walk every new team through. It is deliberately boring. The boring part is what works.
- Define the ICP. One slice of the market, not five. Name the industry, the headcount band, the geography, the job title, and the one problem you will lead with. If you cannot fit that on one line, it is too broad. "VP Product at Series B SaaS companies, 50 to 200 headcount, US and UK, hiring for growth" is a good slice. "B2B buyers" is not.
- Warm up the sender accounts. Fifteen days for new accounts, one week for existing accounts changing pattern. Complete the profile (photo, headline, About, one or two recent posts) before you start. Do not move to the next step until each account is at day 15 of its curve. Full detail is in LinkedIn account warm-up.
- Source leads with Sales Navigator filters. Pull 500 to 2,000 prospects that match the ICP. Use current job title, seniority, headcount, geography, and time-in-role. Eyeball the first 50 results to sanity-check the filters. In Flow AI, add them straight to a list from the Find Leads search; outside Flow, export carefully and respect LinkedIn's terms.
- Draft a first message. Two parts: specific compliment, qualifying question. Keep it under three sentences. Lift the structure from the Outreach Playbook, test it on five manual sends first, then roll it out. Do not pitch in the first message. Do not drop a link. Do not mention your company unless they ask.
- Set cadence and send rules. 15 connection requests per day per account, 9am to 6pm local time, spread across the week. Two scheduled follow-ups a week apart if they do not reply, then stop. Withdraw requests after 21 days. If you are on Flow AI's Autopilot, these settings are the defaults; if not, build them into your own tool's cadence.
- Track the metrics and iterate. Weekly review of acceptance rate, reply rate, positive reply rate, and meetings booked, sliced by sender and by list. Change one variable a week. Keep what works, kill what does not. After 30 days, you will know whether the ICP or the message is the weak link; fix that before you raise volume.
Two meta-rules on top of the six. One, do not add a second campaign until the first one has stable numbers. It is tempting to run three experiments at once; you will learn less from three half-baked campaigns than from one that has two weeks of clean data. Two, treat the system as a living thing. The channel drifts. What worked in January needs a tune-up by April. Build the habit of reading your own metrics before you ever read a new playbook.
If you want the commercial version of this system rather than the reference one, the LinkedIn automation page is where we walk through how Flow AI implements every step in the list above, in one workspace. This guide is the explanation; that page is the pitch. Read this first, then that.
Choosing a LinkedIn automation tool
This section is deliberately short because the page you want for tool selection is not this one. It is our Compare hub, which puts Flow AI next to every major alternative in the category (Dripify, Expandi, HeyReach, Linked Helper, MeetAlfred, and so on) on a feature-by-feature basis.
For this guide, four criteria matter when you shortlist. First, is the automation cloud-based with proper rate limiting, or is it a browser extension you have to keep open? Cloud, every time. Second, does it handle multi-sender natively, with prospect rotation and a shared inbox, or is it a single-seat tool with multi-user billing stapled on? Native, every time. Third, does it have search, warm-up, cadence, and reply management in one place, or do you need three tools glued together? One place, every time. Fourth, does it do AI drafting in your voice with human approval, or does it auto-send? Human approval, every time. Flow AI sits on the correct side of all four by design, which is why we built it. Your mileage on alternatives will vary; use the compare pages to do the homework.
If you would rather see the product answer to each of the six setup steps above, the commercial LinkedIn automation page walks through exactly that. Treat this section as a pointer. The real shopping happens on those two hubs.
Frequently asked questions
Below is a short version of the questions that come up every week in customer calls. Full answers are in the FAQ block at the end of this page.
That is the whole guide. LinkedIn automation in 2026 is not about volume; it is about running a tight system on a handful of accounts, writing messages a human would actually reply to, and measuring the four numbers that matter. Build that, keep it healthy, and the channel becomes one of the most reliable sources of meetings a B2B team can have. If you want the exact message patterns I use, the one resource to read next is the Outreach Playbook, which is the source for every sample DM in this guide and the step-by-step we give every customer on day one.