diff --git a/README.md b/README.md
index fae04f0..32b94a1 100644
--- a/README.md
+++ b/README.md
@@ -21,6 +21,7 @@
 * %first - read first message
 * %rand - read a random message
 * %last - read last message
+* %find - find specific words or phrases
 * %repeat - repeat last analysis (adding supplied arguments)
 * %mobile - fix @invalid-user for last command but mentions users
 * %gdpr - displays GDPR information
@@ -115,6 +116,7 @@ python3 src/main.py
 * **v1.14**
   * `mobile/mention` arg to fix mobile bug
   * `%repeat`, `%mobile` to repeat commands
+  * more scan: `%find`
   * bug fix
 * **v1.13**
   * improved scan `%words`
diff --git a/src/data_types/counter.py b/src/data_types/counter.py
index 5f6d569..212cf89 100644
--- a/src/data_types/counter.py
+++ b/src/data_types/counter.py
@@ -14,7 +14,7 @@ class Counter:
 
     def update_use(self, count: int, date: datetime, item: int = 0):
         self.usages[item] += count
-        if self.last_used is None or date > self.last_used:
+        if count > 0 and (self.last_used is None or date > self.last_used):
             self.last_used = date
 
     def score(self) -> float:
@@ -22,6 +22,8 @@ class Counter:
         # When 2 emojis have the same score,
         # the days since last use is stored in the digits
         # (more recent first)
+        if self.last_used is None:
+            return 0
         return self.all_usages() + 1 / (
             100000 * ((datetime.today() - self.last_used).days + 1)
         )
@@ -59,7 +61,7 @@ class Counter:
         else:
             output += f"{name} - unused"
         top_item = top_key(self.usages)
-        if top and top_item != 0 and transform is not None:
+        if sum > 0 and top and top_item != 0 and transform is not None:
             if self.usages[top_item] == sum:
                 output += f" (all{transform(top_item)})"
             else:
diff --git a/src/main.py b/src/main.py
index 1cee0f6..97dda5b 100644
--- a/src/main.py
+++ b/src/main.py
@@ -68,6 +68,12 @@ bot.register_command(
     "mobile: fix @invalid-user for last command but mentions users",
     "```\n%mobile: fix @invalid-user for last command but mentions users\n```",
 )
+bot.register_command(
+    "find",
+    lambda *args: scanners.FindScanner().compute(*args),
+    "find: find specific words or phrases",
+    scanners.FindScanner.help(),
+)
 bot.register_command(
     "last",
     lambda *args: scanners.LastScanner().compute(*args),
diff --git a/src/scanners/__init__.py b/src/scanners/__init__.py
index aca349b..c174ed7 100644
--- a/src/scanners/__init__.py
+++ b/src/scanners/__init__.py
@@ -3,6 +3,7 @@ from .scanner import Scanner
 from .channels_scanner import ChannelsScanner
 from .composition_scanner import CompositionScanner
 from .emojis_scanner import EmojisScanner
+from .find_scanner import FindScanner
 from .first_scanner import FirstScanner
 from .frequency_scanner import FrequencyScanner
 from .full_scanner import FullScanner
diff --git a/src/scanners/find_scanner.py b/src/scanners/find_scanner.py
new file mode 100644
index 0000000..334beb0
--- /dev/null
+++ b/src/scanners/find_scanner.py
@@ -0,0 +1,99 @@
+from typing import Dict, List
+from collections import defaultdict
+import discord
+import re
+
+# Custom libs
+
+from logs import ChannelLogs, MessageLog
+from .scanner import Scanner
+from data_types import Counter
+from utils import (
+    generate_help,
+    plural,
+    precise,
+    mention,
+)
+
+
+class FindScanner(Scanner):
+    @staticmethod
+    def help() -> str:
+        return generate_help(
+            "find",
+            "Find specific words or phrases",
+            args=[
+                "all/everyone - include bots",
+            ],
+            example='#mychannel1 #mychannel2 @user "I love you" "you too"',
+        )
+
+    def __init__(self):
+        super().__init__(
+            all_args=True,
+            valid_args=["all", "everyone"],
+            help=FindScanner.help(),
+            intro_context="Matches",
+        )
+
+    async def init(self, message: discord.Message, *args: str) -> bool:
+        self.matches = defaultdict(Counter)
+        self.all_messages = "all" in args or "everyone" in args
+        return True
+
+    def compute_message(self, channel: ChannelLogs, message: MessageLog):
+        return FindScanner.analyse_message(
+            message,
+            self.matches,
+            self.other_args,
+            self.raw_members,
+            all_messages=self.all_messages,
+        )
+
+    def get_results(self, intro: str) -> List[str]:
+        matches = [match for match in self.matches]
+        matches.sort(key=lambda match: self.matches[match].score(), reverse=True)
+        usage_count = Counter.total(self.matches)
+        res = [intro]
+        res += [
+            self.matches[match].to_string(
+                matches.index(match),
+                f"`{match}`",
+                total_usage=self.msg_count,
+                ranking=False,
+                transform=lambda id: f" by {mention(id)}",
+                top=len(self.members) != 1,
+            )
+            for match in matches
+        ]
+        if len(matches) > 1:
+            res += [
+                f"Total: {plural(usage_count,'time')} ({precise(usage_count/self.msg_count)}/msg)"
+            ]
+        return res
+
+    special_cases = ["'s", "s"]
+
+    @staticmethod
+    def analyse_message(
+        message: MessageLog,
+        matches: Dict[str, Counter],
+        queries: List[str],
+        raw_members: List[int],
+        *,
+        all_messages: bool,
+    ) -> bool:
+        impacted = False
+        # If author is included in the selection (empty list is all)
+        if (
+            (not message.bot or all_messages)
+            and len(raw_members) == 0
+            or message.author in raw_members
+        ):
+            impacted = True
+            content = message.content.lower()
+            for query in queries:
+                matches[query].update_use(
+                    content.count(query.lower()), message.created_at, message.author
+                )
+        return impacted